diff --git a/Code/MonoMutliViewClassifiers/ExecClassif.py b/Code/MonoMutliViewClassifiers/ExecClassif.py index 34ab48ffc5028b79c395c61e861dbcda9730f4b4..b58cb12905a4f70beb85977f3136632e37bfacaa 100644 --- a/Code/MonoMutliViewClassifiers/ExecClassif.py +++ b/Code/MonoMutliViewClassifiers/ExecClassif.py @@ -311,26 +311,29 @@ else: bestClassifiers.append(classifiersNames[viewIndex][np.argmax(np.array(accuracies[viewIndex]))]) bestClassifiersConfigs.append(classifiersConfigs[viewIndex][np.argmax(np.array(accuracies[viewIndex]))]) monoviewTime = time.time()-dataBaseTime +print resultsMonoview try: if benchmark["Multiview"]: - try: - if benchmark["Multiview"]["Mumbo"]: - for combination in itertools.combinations_with_replacement(range(len(benchmark["Multiview"]["Mumbo"])), NB_VIEW): - classifiersNames = [benchmark["Multiview"]["Mumbo"][index] for index in combination] - arguments = {"CL_type": "Mumbo", - "views": args.views.split(":"), - "NB_VIEW": len(args.views.split(":")), - "NB_CLASS": len(args.CL_classes.split(":")), - "LABELS_NAMES": args.CL_classes.split(":"), - "MumboKWARGS": {"classifiersNames": classifiersNames, - "maxIter":int(args.MU_iter[0]), "minIter":int(args.MU_iter[1]), - "threshold":args.MU_iter[2], - "classifiersConfigs": [argument.split(":") for argument in args.MU_config]}} - argumentDictionaries["Multiview"].append(arguments) - except: - pass -# bestClassifiers = ["DecisionTree", "DecisionTree", "DecisionTree", "DecisionTree"] - + # try: + # if benchmark["Multiview"]["Mumbo"]: + # for combination in itertools.combinations_with_replacement(range(len(benchmark["Multiview"]["Mumbo"])), NB_VIEW): + # classifiersNames = [benchmark["Multiview"]["Mumbo"][index] for index in combination] + # arguments = {"CL_type": "Mumbo", + # "views": args.views.split(":"), + # "NB_VIEW": len(args.views.split(":")), + # "NB_CLASS": len(args.CL_classes.split(":")), + # "LABELS_NAMES": args.CL_classes.split(":"), + # "MumboKWARGS": {"classifiersNames": classifiersNames, + # "maxIter":int(args.MU_iter[0]), "minIter":int(args.MU_iter[1]), + # "threshold":args.MU_iter[2], + # "classifiersConfigs": [argument.split(":") for argument in args.MU_config]}} + # argumentDictionaries["Multiview"].append(arguments) + # except: + # pass + # bestClassifiers = ["DecisionTree", "DecisionTree", "DecisionTree", "DecisionTree"] + # monoviewTime = 0 + # resultsMonoview = [] + # bestClassifiersConfigs = [] try: if benchmark["Multiview"]["Fusion"]: try: @@ -368,6 +371,7 @@ try: pass except: pass +# resultsMultiview = [] if nbCores>1: resultsMultiview = [] nbExperiments = len(argumentDictionaries["Multiview"]) @@ -375,12 +379,12 @@ if nbCores>1: resultsMultiview += Parallel(n_jobs=nbCores)( delayed(ExecMultiview_multicore)(coreIndex, args.name, args.CL_split, args.CL_nbFolds, args.type, args.pathF, LABELS_DICTIONARY, gridSearch=gridSearch, - metrics=metrics, **argumentDictionaries["Multiview"][stepIndex*nbCores+coreIndex]) + metrics=metrics, nIter=args.CL_GS_iter, **argumentDictionaries["Multiview"][stepIndex*nbCores+coreIndex]) for coreIndex in range(min(nbCores, nbExperiments - (stepIndex + 1) * nbCores))) else: resultsMultiview = [ExecMultiview(DATASET, args.name, args.CL_split, args.CL_nbFolds, 1, args.type, args.pathF, LABELS_DICTIONARY, gridSearch=gridSearch, - metrics=metrics, **arguments) for arguments in argumentDictionaries["Multiview"]] + metrics=metrics, nIter=args.CL_GS_iter, **arguments) for arguments in argumentDictionaries["Multiview"]] multiviewTime = time.time()-monoviewTime if nbCores>1: logging.debug("Start:\t Deleting "+str(nbCores)+" temporary datasets for multiprocessing") @@ -390,6 +394,7 @@ if nbCores>1: times = [dataBaseTime, monoviewTime, multiviewTime] # times=[] results = (resultsMonoview, resultsMultiview) +logging.debug("Start:\t Analyze Results") resultAnalysis(benchmark, results, args.name, times, metrics) - +logging.debug("Done:\t Analyze Results") diff --git a/Code/MonoMutliViewClassifiers/Monoview/ExecClassifMonoView.py b/Code/MonoMutliViewClassifiers/Monoview/ExecClassifMonoView.py index 577a4999a8607e7198573f2751156a70b8477699..154a52c621ccb09c203e7c8057725adca999d03f 100644 --- a/Code/MonoMutliViewClassifiers/Monoview/ExecClassifMonoView.py +++ b/Code/MonoMutliViewClassifiers/Monoview/ExecClassifMonoView.py @@ -107,6 +107,7 @@ def ExecMonoview(X, Y, name, learningRate, nbFolds, nbCores, databaseType, path, clKWARGS, classLabelsNames, X.shape, y_train, y_train_pred, y_test, y_test_pred, t_end) cl_desc = [value for key, value in sorted(clKWARGS.iteritems())] + print cl_desc logging.debug("Done:\t Getting Results") logging.info(stringAnalysis) labelsString = "-".join(classLabelsNames) @@ -132,7 +133,7 @@ def ExecMonoview(X, Y, name, learningRate, nbFolds, nbCores, databaseType, path, logging.info("Done:\t Result Analysis") viewIndex = args["viewIndex"] - return viewIndex, [CL_type, cl_desc.append(feat), metricsScores] + return viewIndex, [CL_type, cl_desc+[feat], metricsScores] # # Classification Report with Precision, Recall, F1 , Support # logging.debug("Info:\t Classification report:") # filename = datetime.datetime.now().strftime("%Y_%m_%d") + "-CMV-" + name + "-" + feat + "-Report" diff --git a/Code/MonoMutliViewClassifiers/Multiview/Fusion/Methods/EarlyFusionPackage/WeightedLinear.py b/Code/MonoMutliViewClassifiers/Multiview/Fusion/Methods/EarlyFusionPackage/WeightedLinear.py index 7af341b5e9d233473d1829e8635214cb84b7fb62..61707487d14c014bfb06bc63e7312499dcae4844 100644 --- a/Code/MonoMutliViewClassifiers/Multiview/Fusion/Methods/EarlyFusionPackage/WeightedLinear.py +++ b/Code/MonoMutliViewClassifiers/Multiview/Fusion/Methods/EarlyFusionPackage/WeightedLinear.py @@ -34,12 +34,11 @@ class WeightedLinear(EarlyFusionClassifier): self.weights = self.weights/float(max(self.weights)) self.makeMonoviewData_hdf5(DATASET, weights=self.weights, usedIndices=trainIndices) monoviewClassifierModule = getattr(MonoviewClassifiers, self.monoviewClassifierName) - print self.monoviewClassifiersConfig self.monoviewClassifier = monoviewClassifierModule.fit(self.monoviewData, DATASET.get("labels")[trainIndices], - NB_CORES=self.nbCores, **self.monoviewClassifiersConfig) - #**dict((str(configIndex), config) for configIndex, config in - # enumerate(self.monoviewClassifiersConfig - # ))) + NB_CORES=self.nbCores, #**self.monoviewClassifiersConfig) + **dict((str(configIndex), config) for configIndex, config in + enumerate(self.monoviewClassifiersConfig + ))) def predict_hdf5(self, DATASET, usedIndices=None): self.weights = self.weights/float(max(self.weights)) diff --git a/Code/MonoMutliViewClassifiers/Multiview/Fusion/Methods/LateFusionPackage/WeightedLinear.py b/Code/MonoMutliViewClassifiers/Multiview/Fusion/Methods/LateFusionPackage/WeightedLinear.py index d3e05e42eb6f7d0dbb531878fa8b5bc817871d1d..9bb535f09ba3e60c0a9a96841b57f4bc112c98ff 100644 --- a/Code/MonoMutliViewClassifiers/Multiview/Fusion/Methods/LateFusionPackage/WeightedLinear.py +++ b/Code/MonoMutliViewClassifiers/Multiview/Fusion/Methods/LateFusionPackage/WeightedLinear.py @@ -26,7 +26,10 @@ class WeightedLinear(LateFusionClassifier): def __init__(self, NB_CORES=1, **kwargs): LateFusionClassifier.__init__(self, kwargs['classifiersNames'], kwargs['classifiersConfigs'], NB_CORES=NB_CORES) - self.weights = np.array(map(float, kwargs['fusionMethodConfig'][0])) + if kwargs['fusionMethodConfig'][0]==None: + self.weights = np.ones(len(kwargs["classifiersNames"]), dtype=float) + else: + self.weights = np.array(map(float, kwargs['fusionMethodConfig'][0])) def predict_hdf5(self, DATASET, usedIndices=None): self.weights = self.weights/float(max(self.weights)) diff --git a/Code/MonoMutliViewClassifiers/Multiview/GetMultiviewDb.py b/Code/MonoMutliViewClassifiers/Multiview/GetMultiviewDb.py index b944ffa99b61febd59bce637c34cd459f7d6602f..8a60070a1a6adce323c106ef1f047ea4bfe2bb4f 100644 --- a/Code/MonoMutliViewClassifiers/Multiview/GetMultiviewDb.py +++ b/Code/MonoMutliViewClassifiers/Multiview/GetMultiviewDb.py @@ -1,4 +1,5 @@ import numpy as np +import math from string import digits import os import random @@ -259,6 +260,54 @@ def getMultiOmicDBcsv(features, path, name, NB_CLASS, LABELS_NAMES): return datasetFile, labelDictionary +def findClosestPowerOfTwo(k): + power=1 + while k-power>0: + power = 2*power + if abs(k-power)<abs(k-power/2): + return power + else: + return power/2 + + +def getVector(matrix): + argmax = [0,0] + n = len(matrix) + maxi = 0 + for i in range(n): + for j in range(n): + if j==i+1: + value = (i+1)*(n-j) + if value>maxi: + maxi= value + argmax = [i,j] + i,j = argmax + vector = np.zeros(n, dtype=bool) + vector[:i+1]=np.ones(i+1, dtype=bool) + matrixSup = [i+1, j+1] + matrixInf = [i+1, j+1] + return vector, matrixSup, matrixInf + + +def easyFactorize(targetMatrix, k, t=0): + n = len(targetMatrix) + if math.log(k+1, 2)%1==0.0: + pass + else: + k = findClosestPowerOfTwo(k)-1 + if k==1: + t=1 + return t, getVector(targetMatrix)[0] + vector, matrixSup, matrixInf = getVector(targetMatrix) + t, vectorSup = easyFactorize(targetMatrix[:matrixSup[0], :matrixSup[1]], (k-1)/2, t) + t, vectorInf = easyFactorize(targetMatrix[matrixInf[0]:, matrixInf[0]:], (k-1)/2, t) + factor = np.zeros((n,2*t+1), dtype=bool) + factor[:matrixSup[0], :t] = vectorSup.reshape(factor[:matrixSup[0], :t].shape) + factor[matrixInf[0]:, t:2*t] = vectorInf.reshape(factor[matrixInf[0]:, t:2*t].shape) + factor[:, 2*t] = vector + return 2*t+1, factor + + def getModifiedMultiOmicDBcsv(features, path, name, NB_CLASS, LABELS_NAMES): datasetFile = h5py.File(path+"ModifiedMultiOmic.hdf5", "w") @@ -307,6 +356,16 @@ def getModifiedMultiOmicDBcsv(features, path, name, NB_CLASS, LABELS_NAMES): mrnaseqDset.attrs["name"] = "SRNASeq" logging.debug("Done:\t Getting Sorted RNASeq Data") + logging.debug("Start:\t Getting Binarized RNASeq Data") + factorizedBaseMatrix = np.genfromtxt(path+"factorMatrix.csv", delimiter=',') + brnaseqDset = datasetFile.create_dataset("View5", len(modifiedRNASeq), len(factorizedBaseMatrix.flatten())) + for patientIndex, patientSortedArray in enumerate(modifiedRNASeq): + patientMatrix = np.zeros(factorizedBaseMatrix.shape, dtype=bool) + for lineIndex, geneIndex in enumerate(patientSortedArray): + patientMatrix[geneIndex]=factorizedBaseMatrix[lineIndex] + brnaseqDset[patientIndex] = patientMatrix.flatten() + logging.debug("Done:\t Getting Binarized RNASeq Data") + # logging.debug("Start:\t Getting Binned RNASeq Data") # SRNASeq = datasetFile["View4"][...] # nbBins = 372 diff --git a/Code/MonoMutliViewClassifiers/Multiview/Mumbo/Mumbo.py b/Code/MonoMutliViewClassifiers/Multiview/Mumbo/Mumbo.py index 9a894de3e51f4d6526876c8452926b39b0596d88..7214759ae323465701a72467a60a361d66bb56b7 100644 --- a/Code/MonoMutliViewClassifiers/Multiview/Mumbo/Mumbo.py +++ b/Code/MonoMutliViewClassifiers/Multiview/Mumbo/Mumbo.py @@ -50,7 +50,7 @@ def gridSearch_hdf5(DATASET, classificationKWARGS, learningIndices, metric=None, classifiersNames = classificationKWARGS["classifiersNames"] bestSettings = [] for classifierIndex, classifierName in enumerate(classifiersNames): - logging.debug("\tStart:\t Gridsearch for "+classifierName+" on "+DATASET.get("View"+str(classifierIndex)).attrs["name"]) + logging.debug("\tStart:\t Random search for "+classifierName+" on "+DATASET.get("View"+str(classifierIndex)).attrs["name"]) classifierModule = globals()[classifierName] # Permet d'appeler une fonction avec une string classifierMethod = getattr(classifierModule, "gridSearch") bestSettings.append(classifierMethod(DATASET.get("View"+str(classifierIndex))[learningIndices], diff --git a/Code/MonoMutliViewClassifiers/ResultAnalysis.py b/Code/MonoMutliViewClassifiers/ResultAnalysis.py index ff02817fe307d69623975821d7f52d1e4398a75b..9782f7d57cc3301c59f80c5e1991bc85517eb808 100644 --- a/Code/MonoMutliViewClassifiers/ResultAnalysis.py +++ b/Code/MonoMutliViewClassifiers/ResultAnalysis.py @@ -20,7 +20,9 @@ def resultAnalysis(benchmark, results, name, times, metrics): for metric in metrics: mono, multi = results names = [res[1][0]+"-"+res[1][1][-1] for res in mono] - names+=[type_ if type_ != "Fusion" else a["fusionType"]+"-"+a["fusionMethod"] for type_, a, b in multi] + names+=[type_ for type_, a, b in multi if type_ != "Fusion"] + names+=[ "Late-"+str(a["fusionMethod"]) for type_, a, b in multi if type_ == "Fusion" and a["fusionType"] != "EarlyFusion"] + names+=[ "Early-"+a["fusionMethod"]+"-"+a["classifiersNames"][0] for type_, a, b in multi if type_ == "Fusion" and a["fusionType"] != "LateFusion"] nbResults = len(mono)+len(multi) validationScores = [float(res[1][2][metric[0]][2]) for res in mono] validationScores += [float(scores[metric[0]][2]) for a, b, scores in multi] diff --git a/Code/MonoMutliViewClassifiers/Results/20160907-162651-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-MultiOmic-LOG.log b/Code/MonoMutliViewClassifiers/Results/20160907-162651-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-MultiOmic-LOG.log new file mode 100644 index 0000000000000000000000000000000000000000..7790e9fe4b578b6955d56e73f1cb1383ef61f7ab --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160907-162651-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-MultiOmic-LOG.log @@ -0,0 +1,224 @@ +2016-09-07 16:26:51,762 DEBUG: Start: Creating 2 temporary datasets for multiprocessing +2016-09-07 16:26:51,762 WARNING: WARNING : /!\ This may use a lot of HDD storage space : 0.273145851562 Gbytes /!\ +2016-09-07 16:27:02,716 DEBUG: Start: Creating datasets for multiprocessing +2016-09-07 16:27:02,972 INFO: Start: Finding all available mono- & multiview algorithms +2016-09-07 16:27:04,713 DEBUG: ### Main Programm for Classification MonoView +2016-09-07 16:27:04,714 DEBUG: ### Classification - Database:MultiOmic Feature:Methyl train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-07 16:27:04,714 DEBUG: Start: Determine Train/Test split +2016-09-07 16:27:04,911 DEBUG: Info: Shape X_train:(242, 25978), Length of y_train:242 +2016-09-07 16:27:04,911 DEBUG: Info: Shape X_test:(105, 25978), Length of y_test:105 +2016-09-07 16:27:04,911 DEBUG: Done: Determine Train/Test split +2016-09-07 16:27:04,911 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-07 16:27:06,463 DEBUG: ### Main Programm for Classification MonoView +2016-09-07 16:27:06,464 DEBUG: ### Classification - Database:MultiOmic Feature:Methyl train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-07 16:27:06,464 DEBUG: Start: Determine Train/Test split +2016-09-07 16:27:06,499 DEBUG: Info: Shape X_train:(242, 25978), Length of y_train:242 +2016-09-07 16:27:06,499 DEBUG: Info: Shape X_test:(105, 25978), Length of y_test:105 +2016-09-07 16:27:06,499 DEBUG: Done: Determine Train/Test split +2016-09-07 16:27:06,499 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-07 16:27:14,206 DEBUG: Done: RandomSearch best settings +2016-09-07 16:27:14,206 DEBUG: Start: Training +2016-09-07 16:27:15,153 DEBUG: Done: RandomSearch best settings +2016-09-07 16:27:15,153 DEBUG: Start: Training +2016-09-07 16:27:16,232 DEBUG: Info: Time for Training: 13.1729319096[s] +2016-09-07 16:27:16,232 DEBUG: Done: Training +2016-09-07 16:27:16,233 DEBUG: Start: Predicting +2016-09-07 16:27:16,348 DEBUG: Done: Predicting +2016-09-07 16:27:16,348 DEBUG: Start: Getting Results +2016-09-07 16:27:16,926 DEBUG: Done: Getting Results +2016-09-07 16:27:16,926 INFO: Classification on MultiOmic database for Methyl with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.771428571429 + +Database configuration : + - Database name : MultiOmic + - View name : Methyl View shape : (347, 25978) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 9 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.771428571429 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.586206896552 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.586206896552 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.228571428571 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.771428571429 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.440385506051 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.68 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.515151515152 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.70202020202 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.228571428571 + + + Classification took 0:00:13 +2016-09-07 16:27:16,964 INFO: Done: Result Analysis +2016-09-07 16:27:17,029 DEBUG: Info: Time for Training: 13.9593689442[s] +2016-09-07 16:27:17,029 DEBUG: Done: Training +2016-09-07 16:27:17,029 DEBUG: Start: Predicting +2016-09-07 16:27:17,043 DEBUG: Done: Predicting +2016-09-07 16:27:17,043 DEBUG: Start: Getting Results +2016-09-07 16:27:17,076 DEBUG: Done: Getting Results +2016-09-07 16:27:17,076 INFO: Classification on MultiOmic database for Methyl with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.790476190476 + +Database configuration : + - Database name : MultiOmic + - View name : Methyl View shape : (347, 25978) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 14, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.790476190476 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.592592592593 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.592592592593 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.209523809524 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.790476190476 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.482108339669 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.761904761905 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.484848484848 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.707702020202 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.209523809524 + + + Classification took 0:00:13 +2016-09-07 16:27:17,076 INFO: Done: Result Analysis +2016-09-07 16:27:18,338 DEBUG: ### Main Programm for Classification MonoView +2016-09-07 16:27:18,339 DEBUG: ### Classification - Database:MultiOmic Feature:Methyl train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-07 16:27:18,339 DEBUG: Start: Determine Train/Test split +2016-09-07 16:27:18,379 DEBUG: Info: Shape X_train:(242, 25978), Length of y_train:242 +2016-09-07 16:27:18,379 DEBUG: Info: Shape X_test:(105, 25978), Length of y_test:105 +2016-09-07 16:27:18,379 DEBUG: Done: Determine Train/Test split +2016-09-07 16:27:18,379 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-07 16:27:18,521 DEBUG: ### Main Programm for Classification MonoView +2016-09-07 16:27:18,522 DEBUG: ### Classification - Database:MultiOmic Feature:Methyl train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-07 16:27:18,522 DEBUG: Start: Determine Train/Test split +2016-09-07 16:27:18,570 DEBUG: Info: Shape X_train:(242, 25978), Length of y_train:242 +2016-09-07 16:27:18,571 DEBUG: Info: Shape X_test:(105, 25978), Length of y_test:105 +2016-09-07 16:27:18,571 DEBUG: Done: Determine Train/Test split +2016-09-07 16:27:18,571 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-07 16:27:18,830 DEBUG: Done: RandomSearch best settings +2016-09-07 16:27:18,830 DEBUG: Start: Training +2016-09-07 16:27:18,913 DEBUG: Info: Time for Training: 1.72966194153[s] +2016-09-07 16:27:18,914 DEBUG: Done: Training +2016-09-07 16:27:18,914 DEBUG: Start: Predicting +2016-09-07 16:27:18,924 DEBUG: Done: Predicting +2016-09-07 16:27:18,924 DEBUG: Start: Getting Results +2016-09-07 16:27:19,006 DEBUG: Done: Getting Results +2016-09-07 16:27:19,006 INFO: Classification on MultiOmic database for Methyl with RandomForest + +accuracy_score on train : 0.97520661157 +accuracy_score on test : 0.761904761905 + +Database configuration : + - Database name : MultiOmic + - View name : Methyl View shape : (347, 25978) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 6, max_depth : 9 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.97520661157 + - Score on test : 0.761904761905 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.949152542373 + - Score on test : 0.468085106383 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.949152542373 + - Score on test : 0.468085106383 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0247933884298 + - Score on test : 0.238095238095 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.97520661157 + - Score on test : 0.761904761905 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.932999668908 + - Score on test : 0.398313753408 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.965517241379 + - Score on test : 0.785714285714 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.933333333333 + - Score on test : 0.333333333333 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.961172161172 + - Score on test : 0.645833333333 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0247933884298 + - Score on test : 0.238095238095 + + + Classification took 0:00:01 +2016-09-07 16:27:19,007 INFO: Done: Result Analysis diff --git a/Code/MonoMutliViewClassifiers/Results/20160907-162716Results-DecisionTree-Non-Oui-learnRate0.7-MultiOmic.txt b/Code/MonoMutliViewClassifiers/Results/20160907-162716Results-DecisionTree-Non-Oui-learnRate0.7-MultiOmic.txt new file mode 100644 index 0000000000000000000000000000000000000000..e9719acf0c5fcc94a1c5ad47a4f0609ba08055af --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160907-162716Results-DecisionTree-Non-Oui-learnRate0.7-MultiOmic.txt @@ -0,0 +1,54 @@ +Classification on MultiOmic database for Methyl with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.771428571429 + +Database configuration : + - Database name : MultiOmic + - View name : Methyl View shape : (347, 25978) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 9 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.771428571429 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.586206896552 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.586206896552 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.228571428571 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.771428571429 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.440385506051 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.68 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.515151515152 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.70202020202 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.228571428571 + + + Classification took 0:00:13 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160907-162717Results-Adaboost-Non-Oui-learnRate0.7-MultiOmic.txt b/Code/MonoMutliViewClassifiers/Results/20160907-162717Results-Adaboost-Non-Oui-learnRate0.7-MultiOmic.txt new file mode 100644 index 0000000000000000000000000000000000000000..32fa654fd647c09fec791c94b3f3b876d76473b9 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160907-162717Results-Adaboost-Non-Oui-learnRate0.7-MultiOmic.txt @@ -0,0 +1,57 @@ +Classification on MultiOmic database for Methyl with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.790476190476 + +Database configuration : + - Database name : MultiOmic + - View name : Methyl View shape : (347, 25978) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 14, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.790476190476 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.592592592593 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.592592592593 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.209523809524 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.790476190476 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.482108339669 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.761904761905 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.484848484848 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.707702020202 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.209523809524 + + + Classification took 0:00:13 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160907-162719Results-RandomForest-Non-Oui-learnRate0.7-MultiOmic.txt b/Code/MonoMutliViewClassifiers/Results/20160907-162719Results-RandomForest-Non-Oui-learnRate0.7-MultiOmic.txt new file mode 100644 index 0000000000000000000000000000000000000000..97c69cd4414a34cedd3472bb2f1a24d98437e90b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160907-162719Results-RandomForest-Non-Oui-learnRate0.7-MultiOmic.txt @@ -0,0 +1,54 @@ +Classification on MultiOmic database for Methyl with RandomForest + +accuracy_score on train : 0.97520661157 +accuracy_score on test : 0.761904761905 + +Database configuration : + - Database name : MultiOmic + - View name : Methyl View shape : (347, 25978) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 6, max_depth : 9 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.97520661157 + - Score on test : 0.761904761905 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.949152542373 + - Score on test : 0.468085106383 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.949152542373 + - Score on test : 0.468085106383 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0247933884298 + - Score on test : 0.238095238095 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.97520661157 + - Score on test : 0.761904761905 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.932999668908 + - Score on test : 0.398313753408 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.965517241379 + - Score on test : 0.785714285714 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.933333333333 + - Score on test : 0.333333333333 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.961172161172 + - Score on test : 0.645833333333 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0247933884298 + - Score on test : 0.238095238095 + + + Classification took 0:00:01 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095527-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log b/Code/MonoMutliViewClassifiers/Results/20160908-095527-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log new file mode 100644 index 0000000000000000000000000000000000000000..52d4e73fa353f4c5e4dd770a4fbf8026a0dfdd7b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095527-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log @@ -0,0 +1,2211 @@ +2016-09-08 09:55:28,079 DEBUG: Start: Creating 2 temporary datasets for multiprocessing +2016-09-08 09:55:28,079 WARNING: WARNING : /!\ This may use a lot of HDD storage space : 0.00010290625 Gbytes /!\ +2016-09-08 09:55:33,093 DEBUG: Start: Creating datasets for multiprocessing +2016-09-08 09:55:33,096 INFO: Start: Finding all available mono- & multiview algorithms +2016-09-08 09:55:33,343 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:33,343 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:33,343 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:55:33,343 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:55:33,343 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:33,343 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:33,383 DEBUG: Info: Shape X_train:(210, 8), Length of y_train:210 +2016-09-08 09:55:33,383 DEBUG: Info: Shape X_train:(210, 8), Length of y_train:210 +2016-09-08 09:55:33,383 DEBUG: Info: Shape X_test:(90, 8), Length of y_test:90 +2016-09-08 09:55:33,384 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:33,384 DEBUG: Info: Shape X_test:(90, 8), Length of y_test:90 +2016-09-08 09:55:33,384 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:33,384 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:33,384 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:33,415 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:33,415 DEBUG: Start: Training +2016-09-08 09:55:33,417 DEBUG: Info: Time for Training: 0.0742340087891[s] +2016-09-08 09:55:33,417 DEBUG: Done: Training +2016-09-08 09:55:33,417 DEBUG: Start: Predicting +2016-09-08 09:55:33,439 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:33,440 DEBUG: Start: Training +2016-09-08 09:55:33,445 DEBUG: Info: Time for Training: 0.102918863297[s] +2016-09-08 09:55:33,446 DEBUG: Done: Training +2016-09-08 09:55:33,446 DEBUG: Start: Predicting +2016-09-08 09:55:33,583 DEBUG: Done: Predicting +2016-09-08 09:55:33,583 DEBUG: Done: Predicting +2016-09-08 09:55:33,584 DEBUG: Start: Getting Results +2016-09-08 09:55:33,584 DEBUG: Start: Getting Results +2016-09-08 09:55:34,228 DEBUG: Done: Getting Results +2016-09-08 09:55:34,228 INFO: Classification on Fake database for View0 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 4, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.505263157895 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.505263157895 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0418655345164 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.470588235294 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.545454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.479249011858 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 +2016-09-08 09:55:34,228 DEBUG: Done: Getting Results +2016-09-08 09:55:34,228 INFO: Classification on Fake database for View0 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.484210526316 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.484210526316 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0867214643554 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.450980392157 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522727272727 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.457015810277 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:55:34,228 INFO: Done: Result Analysis +2016-09-08 09:55:34,228 INFO: Done: Result Analysis +2016-09-08 09:55:34,286 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:34,286 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:34,287 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 09:55:34,287 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 09:55:34,287 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:34,287 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:34,287 DEBUG: Info: Shape X_train:(210, 8), Length of y_train:210 +2016-09-08 09:55:34,287 DEBUG: Info: Shape X_train:(210, 8), Length of y_train:210 +2016-09-08 09:55:34,288 DEBUG: Info: Shape X_test:(90, 8), Length of y_test:90 +2016-09-08 09:55:34,288 DEBUG: Info: Shape X_test:(90, 8), Length of y_test:90 +2016-09-08 09:55:34,288 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:34,288 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:34,288 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:34,288 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:34,320 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:34,320 DEBUG: Start: Training +2016-09-08 09:55:34,321 DEBUG: Info: Time for Training: 0.0349078178406[s] +2016-09-08 09:55:34,321 DEBUG: Done: Training +2016-09-08 09:55:34,321 DEBUG: Start: Predicting +2016-09-08 09:55:34,326 DEBUG: Done: Predicting +2016-09-08 09:55:34,326 DEBUG: Start: Getting Results +2016-09-08 09:55:34,371 DEBUG: Done: Getting Results +2016-09-08 09:55:34,371 INFO: Classification on Fake database for View0 with KNN + +accuracy_score on train : 0.661904761905 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.661904761905 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.727969348659 + - Score on test : 0.534653465347 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.727969348659 + - Score on test : 0.534653465347 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.338095238095 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.661904761905 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.290672377783 + - Score on test : -0.0399755963154 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.68345323741 + - Score on test : 0.473684210526 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.77868852459 + - Score on test : 0.613636363636 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.639344262295 + - Score on test : 0.480731225296 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.338095238095 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 +2016-09-08 09:55:34,371 INFO: Done: Result Analysis +2016-09-08 09:55:34,386 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:34,387 DEBUG: Start: Training +2016-09-08 09:55:34,397 DEBUG: Info: Time for Training: 0.11102604866[s] +2016-09-08 09:55:34,397 DEBUG: Done: Training +2016-09-08 09:55:34,397 DEBUG: Start: Predicting +2016-09-08 09:55:34,400 DEBUG: Done: Predicting +2016-09-08 09:55:34,401 DEBUG: Start: Getting Results +2016-09-08 09:55:34,429 DEBUG: Done: Getting Results +2016-09-08 09:55:34,429 INFO: Classification on Fake database for View0 with RandomForest + +accuracy_score on train : 0.895238095238 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 4, max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.895238095238 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.905982905983 + - Score on test : 0.515463917526 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.905982905983 + - Score on test : 0.515463917526 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.104761904762 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.895238095238 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.791868570857 + - Score on test : -0.0411594726194 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.946428571429 + - Score on test : 0.471698113208 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.868852459016 + - Score on test : 0.568181818182 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.900335320417 + - Score on test : 0.479743083004 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.104761904762 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 +2016-09-08 09:55:34,430 INFO: Done: Result Analysis +2016-09-08 09:55:34,537 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:34,537 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 09:55:34,537 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:34,537 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:34,538 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 09:55:34,538 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:34,538 DEBUG: Info: Shape X_train:(210, 8), Length of y_train:210 +2016-09-08 09:55:34,538 DEBUG: Info: Shape X_test:(90, 8), Length of y_test:90 +2016-09-08 09:55:34,538 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:34,538 DEBUG: Info: Shape X_train:(210, 8), Length of y_train:210 +2016-09-08 09:55:34,538 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:34,538 DEBUG: Info: Shape X_test:(90, 8), Length of y_test:90 +2016-09-08 09:55:34,538 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:34,539 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:34,668 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:34,668 DEBUG: Start: Training +2016-09-08 09:55:34,686 DEBUG: Info: Time for Training: 0.148998022079[s] +2016-09-08 09:55:34,686 DEBUG: Done: Training +2016-09-08 09:55:34,686 DEBUG: Start: Predicting +2016-09-08 09:55:34,689 DEBUG: Done: Predicting +2016-09-08 09:55:34,689 DEBUG: Start: Getting Results +2016-09-08 09:55:34,699 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:34,699 DEBUG: Start: Training +2016-09-08 09:55:34,700 DEBUG: Info: Time for Training: 0.164316892624[s] +2016-09-08 09:55:34,700 DEBUG: Done: Training +2016-09-08 09:55:34,700 DEBUG: Start: Predicting +2016-09-08 09:55:34,711 DEBUG: Done: Predicting +2016-09-08 09:55:34,711 DEBUG: Start: Getting Results +2016-09-08 09:55:34,727 DEBUG: Done: Getting Results +2016-09-08 09:55:34,727 INFO: Classification on Fake database for View0 with SVMLinear + +accuracy_score on train : 0.490476190476 +accuracy_score on test : 0.377777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.490476190476 + - Score on test : 0.377777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.563265306122 + - Score on test : 0.416666666667 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.563265306122 + - Score on test : 0.416666666667 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.509523809524 + - Score on test : 0.622222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.490476190476 + - Score on test : 0.377777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0481411286791 + - Score on test : -0.244017569898 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.560975609756 + - Score on test : 0.384615384615 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.565573770492 + - Score on test : 0.454545454545 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.475968703428 + - Score on test : 0.379446640316 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.509523809524 + - Score on test : 0.622222222222 + + + Classification took 0:00:00 +2016-09-08 09:55:34,727 INFO: Done: Result Analysis +2016-09-08 09:55:34,737 DEBUG: Done: Getting Results +2016-09-08 09:55:34,737 INFO: Classification on Fake database for View0 with SGD + +accuracy_score on train : 0.614285714286 +accuracy_score on test : 0.466666666667 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : elasticnet + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.614285714286 + - Score on test : 0.466666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.749226006192 + - Score on test : 0.630769230769 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.749226006192 + - Score on test : 0.630769230769 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.385714285714 + - Score on test : 0.533333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.614285714286 + - Score on test : 0.466666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.201498784613 + - Score on test : -0.112653159931 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.601990049751 + - Score on test : 0.476744186047 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.991803278689 + - Score on test : 0.931818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.541356184799 + - Score on test : 0.476778656126 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.385714285714 + - Score on test : 0.533333333333 + + + Classification took 0:00:00 +2016-09-08 09:55:34,738 INFO: Done: Result Analysis +2016-09-08 09:55:34,880 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:34,881 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:34,881 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 09:55:34,881 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 09:55:34,881 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:34,881 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:34,881 DEBUG: Info: Shape X_train:(210, 8), Length of y_train:210 +2016-09-08 09:55:34,881 DEBUG: Info: Shape X_train:(210, 8), Length of y_train:210 +2016-09-08 09:55:34,882 DEBUG: Info: Shape X_test:(90, 8), Length of y_test:90 +2016-09-08 09:55:34,882 DEBUG: Info: Shape X_test:(90, 8), Length of y_test:90 +2016-09-08 09:55:34,882 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:34,882 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:34,882 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:34,882 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:34,928 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:34,928 DEBUG: Start: Training +2016-09-08 09:55:34,928 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:34,929 DEBUG: Start: Training +2016-09-08 09:55:34,944 DEBUG: Info: Time for Training: 0.0635361671448[s] +2016-09-08 09:55:34,944 DEBUG: Done: Training +2016-09-08 09:55:34,944 DEBUG: Start: Predicting +2016-09-08 09:55:34,945 DEBUG: Info: Time for Training: 0.0647149085999[s] +2016-09-08 09:55:34,945 DEBUG: Done: Training +2016-09-08 09:55:34,945 DEBUG: Start: Predicting +2016-09-08 09:55:34,947 DEBUG: Done: Predicting +2016-09-08 09:55:34,947 DEBUG: Start: Getting Results +2016-09-08 09:55:34,950 DEBUG: Done: Predicting +2016-09-08 09:55:34,950 DEBUG: Start: Getting Results +2016-09-08 09:55:34,979 DEBUG: Done: Getting Results +2016-09-08 09:55:34,979 INFO: Classification on Fake database for View0 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.468085106383 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.468085106383 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.109345881217 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.44 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.445652173913 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 +2016-09-08 09:55:34,980 INFO: Done: Result Analysis +2016-09-08 09:55:34,995 DEBUG: Done: Getting Results +2016-09-08 09:55:34,995 INFO: Classification on Fake database for View0 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.616822429907 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.616822429907 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.106710653456 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.52380952381 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.75 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.548913043478 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 +2016-09-08 09:55:34,995 INFO: Done: Result Analysis +2016-09-08 09:55:35,123 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:35,123 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:35,123 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:55:35,123 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:55:35,123 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:35,123 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:35,124 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 09:55:35,125 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 09:55:35,125 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 09:55:35,125 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:35,125 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 09:55:35,125 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:35,125 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:35,125 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:35,163 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:35,163 DEBUG: Start: Training +2016-09-08 09:55:35,165 DEBUG: Info: Time for Training: 0.0432438850403[s] +2016-09-08 09:55:35,165 DEBUG: Done: Training +2016-09-08 09:55:35,165 DEBUG: Start: Predicting +2016-09-08 09:55:35,168 DEBUG: Done: Predicting +2016-09-08 09:55:35,168 DEBUG: Start: Getting Results +2016-09-08 09:55:35,177 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:35,177 DEBUG: Start: Training +2016-09-08 09:55:35,181 DEBUG: Info: Time for Training: 0.0592088699341[s] +2016-09-08 09:55:35,181 DEBUG: Done: Training +2016-09-08 09:55:35,181 DEBUG: Start: Predicting +2016-09-08 09:55:35,184 DEBUG: Done: Predicting +2016-09-08 09:55:35,184 DEBUG: Start: Getting Results +2016-09-08 09:55:35,215 DEBUG: Done: Getting Results +2016-09-08 09:55:35,215 INFO: Classification on Fake database for View1 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.48 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.48 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.154858431981 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.428571428571 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.545454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.424901185771 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 +2016-09-08 09:55:35,216 INFO: Done: Result Analysis +2016-09-08 09:55:35,220 DEBUG: Done: Getting Results +2016-09-08 09:55:35,220 INFO: Classification on Fake database for View1 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 4, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.48 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.48 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.154858431981 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.428571428571 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.545454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.424901185771 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 +2016-09-08 09:55:35,220 INFO: Done: Result Analysis +2016-09-08 09:55:35,370 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:35,370 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 09:55:35,370 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:35,370 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:35,370 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 09:55:35,370 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:35,371 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 09:55:35,371 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 09:55:35,371 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 09:55:35,371 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:35,371 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 09:55:35,371 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:35,371 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:35,371 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:35,402 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:35,403 DEBUG: Start: Training +2016-09-08 09:55:35,403 DEBUG: Info: Time for Training: 0.0339629650116[s] +2016-09-08 09:55:35,403 DEBUG: Done: Training +2016-09-08 09:55:35,403 DEBUG: Start: Predicting +2016-09-08 09:55:35,409 DEBUG: Done: Predicting +2016-09-08 09:55:35,409 DEBUG: Start: Getting Results +2016-09-08 09:55:35,450 DEBUG: Done: Getting Results +2016-09-08 09:55:35,450 INFO: Classification on Fake database for View1 with KNN + +accuracy_score on train : 0.571428571429 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.6484375 + - Score on test : 0.541666666667 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.6484375 + - Score on test : 0.541666666667 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.428571428571 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.103477711187 + - Score on test : 0.0260018722022 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.619402985075 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.680327868852 + - Score on test : 0.590909090909 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.550391207154 + - Score on test : 0.512845849802 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.428571428571 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 09:55:35,450 INFO: Done: Result Analysis +2016-09-08 09:55:35,480 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:35,481 DEBUG: Start: Training +2016-09-08 09:55:35,491 DEBUG: Info: Time for Training: 0.121489048004[s] +2016-09-08 09:55:35,491 DEBUG: Done: Training +2016-09-08 09:55:35,491 DEBUG: Start: Predicting +2016-09-08 09:55:35,495 DEBUG: Done: Predicting +2016-09-08 09:55:35,495 DEBUG: Start: Getting Results +2016-09-08 09:55:35,527 DEBUG: Done: Getting Results +2016-09-08 09:55:35,527 INFO: Classification on Fake database for View1 with RandomForest + +accuracy_score on train : 0.9 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 4, max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.9 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.909090909091 + - Score on test : 0.356164383562 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.909090909091 + - Score on test : 0.356164383562 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.1 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.9 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.805030105216 + - Score on test : -0.0560191732057 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.963302752294 + - Score on test : 0.448275862069 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.860655737705 + - Score on test : 0.295454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.907600596125 + - Score on test : 0.473814229249 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.1 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 +2016-09-08 09:55:35,528 INFO: Done: Result Analysis +2016-09-08 09:55:35,610 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:35,610 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:35,610 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 09:55:35,610 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 09:55:35,610 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:35,610 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:35,611 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 09:55:35,611 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 09:55:35,611 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 09:55:35,611 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 09:55:35,611 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:35,611 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:35,611 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:35,611 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:35,656 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:35,656 DEBUG: Start: Training +2016-09-08 09:55:35,657 DEBUG: Info: Time for Training: 0.0472548007965[s] +2016-09-08 09:55:35,657 DEBUG: Done: Training +2016-09-08 09:55:35,657 DEBUG: Start: Predicting +2016-09-08 09:55:35,660 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:35,661 DEBUG: Start: Training +2016-09-08 09:55:35,680 DEBUG: Done: Predicting +2016-09-08 09:55:35,680 DEBUG: Start: Getting Results +2016-09-08 09:55:35,681 DEBUG: Info: Time for Training: 0.0712029933929[s] +2016-09-08 09:55:35,681 DEBUG: Done: Training +2016-09-08 09:55:35,681 DEBUG: Start: Predicting +2016-09-08 09:55:35,684 DEBUG: Done: Predicting +2016-09-08 09:55:35,684 DEBUG: Start: Getting Results +2016-09-08 09:55:35,704 DEBUG: Done: Getting Results +2016-09-08 09:55:35,704 INFO: Classification on Fake database for View1 with SGD + +accuracy_score on train : 0.609523809524 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : elasticnet + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.609523809524 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.745341614907 + - Score on test : 0.661417322835 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.745341614907 + - Score on test : 0.661417322835 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.390476190476 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.609523809524 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.172644893682 + - Score on test : 0.118036588599 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.6 + - Score on test : 0.506024096386 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.983606557377 + - Score on test : 0.954545454545 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.537257824143 + - Score on test : 0.53162055336 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.390476190476 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 +2016-09-08 09:55:35,704 INFO: Done: Result Analysis +2016-09-08 09:55:35,714 DEBUG: Done: Getting Results +2016-09-08 09:55:35,715 INFO: Classification on Fake database for View1 with SVMLinear + +accuracy_score on train : 0.542857142857 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.542857142857 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.606557377049 + - Score on test : 0.584905660377 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.606557377049 + - Score on test : 0.584905660377 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.457142857143 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.542857142857 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0611028315946 + - Score on test : 0.0330758927464 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.606557377049 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.606557377049 + - Score on test : 0.704545454545 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.530551415797 + - Score on test : 0.515316205534 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.457142857143 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 09:55:35,715 INFO: Done: Result Analysis +2016-09-08 09:55:35,860 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:35,860 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:35,860 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 09:55:35,860 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 09:55:35,860 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:35,860 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:35,861 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 09:55:35,861 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 09:55:35,861 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 09:55:35,861 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 09:55:35,861 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:35,861 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:35,861 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:35,861 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:35,908 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:35,908 DEBUG: Start: Training +2016-09-08 09:55:35,910 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:35,910 DEBUG: Start: Training +2016-09-08 09:55:35,925 DEBUG: Info: Time for Training: 0.0658419132233[s] +2016-09-08 09:55:35,925 DEBUG: Done: Training +2016-09-08 09:55:35,925 DEBUG: Start: Predicting +2016-09-08 09:55:35,926 DEBUG: Info: Time for Training: 0.06693816185[s] +2016-09-08 09:55:35,926 DEBUG: Done: Training +2016-09-08 09:55:35,926 DEBUG: Start: Predicting +2016-09-08 09:55:35,929 DEBUG: Done: Predicting +2016-09-08 09:55:35,929 DEBUG: Start: Getting Results +2016-09-08 09:55:35,932 DEBUG: Done: Predicting +2016-09-08 09:55:35,932 DEBUG: Start: Getting Results +2016-09-08 09:55:35,962 DEBUG: Done: Getting Results +2016-09-08 09:55:35,962 INFO: Classification on Fake database for View1 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.358208955224 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.358208955224 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0385036888617 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.521739130435 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.272727272727 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.516798418972 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 +2016-09-08 09:55:35,962 INFO: Done: Result Analysis +2016-09-08 09:55:35,967 DEBUG: Done: Getting Results +2016-09-08 09:55:35,967 INFO: Classification on Fake database for View1 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.645161290323 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.645161290323 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0628694613462 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.909090909091 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.51976284585 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 09:55:35,967 INFO: Done: Result Analysis +2016-09-08 09:55:36,109 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:36,109 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:55:36,110 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:36,110 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:36,110 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:55:36,111 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:36,111 DEBUG: Info: Shape X_train:(210, 7), Length of y_train:210 +2016-09-08 09:55:36,111 DEBUG: Info: Shape X_test:(90, 7), Length of y_test:90 +2016-09-08 09:55:36,111 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:36,111 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:36,112 DEBUG: Info: Shape X_train:(210, 7), Length of y_train:210 +2016-09-08 09:55:36,112 DEBUG: Info: Shape X_test:(90, 7), Length of y_test:90 +2016-09-08 09:55:36,112 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:36,112 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:36,145 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:36,145 DEBUG: Start: Training +2016-09-08 09:55:36,146 DEBUG: Info: Time for Training: 0.0371689796448[s] +2016-09-08 09:55:36,146 DEBUG: Done: Training +2016-09-08 09:55:36,146 DEBUG: Start: Predicting +2016-09-08 09:55:36,149 DEBUG: Done: Predicting +2016-09-08 09:55:36,149 DEBUG: Start: Getting Results +2016-09-08 09:55:36,159 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:36,159 DEBUG: Start: Training +2016-09-08 09:55:36,163 DEBUG: Info: Time for Training: 0.0549581050873[s] +2016-09-08 09:55:36,163 DEBUG: Done: Training +2016-09-08 09:55:36,163 DEBUG: Start: Predicting +2016-09-08 09:55:36,166 DEBUG: Done: Predicting +2016-09-08 09:55:36,166 DEBUG: Start: Getting Results +2016-09-08 09:55:36,197 DEBUG: Done: Getting Results +2016-09-08 09:55:36,197 INFO: Classification on Fake database for View2 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.466666666667 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.441860465116 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.441860465116 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.533333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.068316965625 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.452380952381 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.431818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.465909090909 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.533333333333 + + + Classification took 0:00:00 +2016-09-08 09:55:36,197 INFO: Done: Result Analysis +2016-09-08 09:55:36,208 DEBUG: Done: Getting Results +2016-09-08 09:55:36,208 INFO: Classification on Fake database for View2 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 4, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.436781609195 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.436781609195 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0899876638096 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.441860465116 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.431818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455039525692 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:55:36,209 INFO: Done: Result Analysis +2016-09-08 09:55:36,356 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:36,356 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 09:55:36,356 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:36,356 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:36,357 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 09:55:36,357 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:36,357 DEBUG: Info: Shape X_train:(210, 7), Length of y_train:210 +2016-09-08 09:55:36,357 DEBUG: Info: Shape X_test:(90, 7), Length of y_test:90 +2016-09-08 09:55:36,357 DEBUG: Info: Shape X_train:(210, 7), Length of y_train:210 +2016-09-08 09:55:36,358 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:36,358 DEBUG: Info: Shape X_test:(90, 7), Length of y_test:90 +2016-09-08 09:55:36,358 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:36,358 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:36,358 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:36,387 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:36,388 DEBUG: Start: Training +2016-09-08 09:55:36,388 DEBUG: Info: Time for Training: 0.0327939987183[s] +2016-09-08 09:55:36,388 DEBUG: Done: Training +2016-09-08 09:55:36,388 DEBUG: Start: Predicting +2016-09-08 09:55:36,394 DEBUG: Done: Predicting +2016-09-08 09:55:36,394 DEBUG: Start: Getting Results +2016-09-08 09:55:36,437 DEBUG: Done: Getting Results +2016-09-08 09:55:36,437 INFO: Classification on Fake database for View2 with KNN + +accuracy_score on train : 0.619047619048 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.619047619048 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.705882352941 + - Score on test : 0.619469026549 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.705882352941 + - Score on test : 0.619469026549 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.380952380952 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.619047619048 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.189221481343 + - Score on test : 0.0665679839847 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.64 + - Score on test : 0.507246376812 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.786885245902 + - Score on test : 0.795454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.586624441133 + - Score on test : 0.528162055336 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.380952380952 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 +2016-09-08 09:55:36,437 INFO: Done: Result Analysis +2016-09-08 09:55:36,463 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:36,463 DEBUG: Start: Training +2016-09-08 09:55:36,473 DEBUG: Info: Time for Training: 0.11700296402[s] +2016-09-08 09:55:36,473 DEBUG: Done: Training +2016-09-08 09:55:36,473 DEBUG: Start: Predicting +2016-09-08 09:55:36,477 DEBUG: Done: Predicting +2016-09-08 09:55:36,477 DEBUG: Start: Getting Results +2016-09-08 09:55:36,509 DEBUG: Done: Getting Results +2016-09-08 09:55:36,509 INFO: Classification on Fake database for View2 with RandomForest + +accuracy_score on train : 0.890476190476 +accuracy_score on test : 0.4 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 4, max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.890476190476 + - Score on test : 0.4 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.904564315353 + - Score on test : 0.357142857143 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.904564315353 + - Score on test : 0.357142857143 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.109523809524 + - Score on test : 0.6 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.890476190476 + - Score on test : 0.4 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.77645053118 + - Score on test : -0.20378096045 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.915966386555 + - Score on test : 0.375 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.893442622951 + - Score on test : 0.340909090909 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.889903129657 + - Score on test : 0.39871541502 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.109523809524 + - Score on test : 0.6 + + + Classification took 0:00:00 +2016-09-08 09:55:36,509 INFO: Done: Result Analysis +2016-09-08 09:55:36,597 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:36,597 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 09:55:36,597 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:36,597 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:36,598 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 09:55:36,598 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:36,598 DEBUG: Info: Shape X_train:(210, 7), Length of y_train:210 +2016-09-08 09:55:36,598 DEBUG: Info: Shape X_test:(90, 7), Length of y_test:90 +2016-09-08 09:55:36,598 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:36,598 DEBUG: Info: Shape X_train:(210, 7), Length of y_train:210 +2016-09-08 09:55:36,598 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:36,598 DEBUG: Info: Shape X_test:(90, 7), Length of y_test:90 +2016-09-08 09:55:36,599 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:36,599 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:36,645 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:36,645 DEBUG: Start: Training +2016-09-08 09:55:36,646 DEBUG: Info: Time for Training: 0.0489931106567[s] +2016-09-08 09:55:36,646 DEBUG: Done: Training +2016-09-08 09:55:36,646 DEBUG: Start: Predicting +2016-09-08 09:55:36,651 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:36,651 DEBUG: Start: Training +2016-09-08 09:55:36,658 DEBUG: Done: Predicting +2016-09-08 09:55:36,658 DEBUG: Start: Getting Results +2016-09-08 09:55:36,670 DEBUG: Info: Time for Training: 0.0731010437012[s] +2016-09-08 09:55:36,670 DEBUG: Done: Training +2016-09-08 09:55:36,670 DEBUG: Start: Predicting +2016-09-08 09:55:36,674 DEBUG: Done: Predicting +2016-09-08 09:55:36,674 DEBUG: Start: Getting Results +2016-09-08 09:55:36,682 DEBUG: Done: Getting Results +2016-09-08 09:55:36,683 INFO: Classification on Fake database for View2 with SGD + +accuracy_score on train : 0.590476190476 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : elasticnet + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.715231788079 + - Score on test : 0.676923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.715231788079 + - Score on test : 0.676923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0945615027077 + - Score on test : 0.210925065403 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.6 + - Score on test : 0.511627906977 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.885245901639 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.533532041729 + - Score on test : 0.54347826087 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 +2016-09-08 09:55:36,683 INFO: Done: Result Analysis +2016-09-08 09:55:36,702 DEBUG: Done: Getting Results +2016-09-08 09:55:36,703 INFO: Classification on Fake database for View2 with SVMLinear + +accuracy_score on train : 0.533333333333 +accuracy_score on test : 0.433333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.433333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.604838709677 + - Score on test : 0.484848484848 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.604838709677 + - Score on test : 0.484848484848 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.566666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.433333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0354605635154 + - Score on test : -0.131720304791 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.595238095238 + - Score on test : 0.436363636364 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.614754098361 + - Score on test : 0.545454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.517604321908 + - Score on test : 0.435770750988 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.566666666667 + + + Classification took 0:00:00 +2016-09-08 09:55:36,703 INFO: Done: Result Analysis +2016-09-08 09:55:36,843 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:36,843 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 09:55:36,843 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:36,843 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:36,843 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 09:55:36,843 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:36,843 DEBUG: Info: Shape X_train:(210, 7), Length of y_train:210 +2016-09-08 09:55:36,844 DEBUG: Info: Shape X_train:(210, 7), Length of y_train:210 +2016-09-08 09:55:36,844 DEBUG: Info: Shape X_test:(90, 7), Length of y_test:90 +2016-09-08 09:55:36,844 DEBUG: Info: Shape X_test:(90, 7), Length of y_test:90 +2016-09-08 09:55:36,844 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:36,844 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:36,844 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:36,844 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:36,891 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:36,891 DEBUG: Start: Training +2016-09-08 09:55:36,892 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:36,892 DEBUG: Start: Training +2016-09-08 09:55:36,907 DEBUG: Info: Time for Training: 0.0647799968719[s] +2016-09-08 09:55:36,907 DEBUG: Done: Training +2016-09-08 09:55:36,907 DEBUG: Start: Predicting +2016-09-08 09:55:36,908 DEBUG: Info: Time for Training: 0.0657360553741[s] +2016-09-08 09:55:36,908 DEBUG: Done: Training +2016-09-08 09:55:36,908 DEBUG: Start: Predicting +2016-09-08 09:55:36,910 DEBUG: Done: Predicting +2016-09-08 09:55:36,910 DEBUG: Start: Getting Results +2016-09-08 09:55:36,913 DEBUG: Done: Predicting +2016-09-08 09:55:36,913 DEBUG: Start: Getting Results +2016-09-08 09:55:36,946 DEBUG: Done: Getting Results +2016-09-08 09:55:36,946 INFO: Classification on Fake database for View2 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.4 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.357142857143 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.357142857143 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.6 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.20378096045 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.375 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.340909090909 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.39871541502 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.6 + + + Classification took 0:00:00 +2016-09-08 09:55:36,946 INFO: Done: Result Analysis +2016-09-08 09:55:36,952 DEBUG: Done: Getting Results +2016-09-08 09:55:36,952 INFO: Classification on Fake database for View2 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.466666666667 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.538461538462 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.538461538462 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.533333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0628694613462 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.636363636364 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.470355731225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.533333333333 + + + Classification took 0:00:00 +2016-09-08 09:55:36,952 INFO: Done: Result Analysis +2016-09-08 09:55:37,093 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:37,093 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:37,093 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:55:37,093 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:37,093 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:55:37,093 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:37,093 DEBUG: Info: Shape X_train:(210, 6), Length of y_train:210 +2016-09-08 09:55:37,093 DEBUG: Info: Shape X_train:(210, 6), Length of y_train:210 +2016-09-08 09:55:37,094 DEBUG: Info: Shape X_test:(90, 6), Length of y_test:90 +2016-09-08 09:55:37,094 DEBUG: Info: Shape X_test:(90, 6), Length of y_test:90 +2016-09-08 09:55:37,094 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:37,094 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:37,094 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:37,094 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:37,124 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:37,124 DEBUG: Start: Training +2016-09-08 09:55:37,125 DEBUG: Info: Time for Training: 0.0331890583038[s] +2016-09-08 09:55:37,126 DEBUG: Done: Training +2016-09-08 09:55:37,126 DEBUG: Start: Predicting +2016-09-08 09:55:37,128 DEBUG: Done: Predicting +2016-09-08 09:55:37,128 DEBUG: Start: Getting Results +2016-09-08 09:55:37,139 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:37,139 DEBUG: Start: Training +2016-09-08 09:55:37,142 DEBUG: Info: Time for Training: 0.0498540401459[s] +2016-09-08 09:55:37,142 DEBUG: Done: Training +2016-09-08 09:55:37,142 DEBUG: Start: Predicting +2016-09-08 09:55:37,145 DEBUG: Done: Predicting +2016-09-08 09:55:37,145 DEBUG: Start: Getting Results +2016-09-08 09:55:37,172 DEBUG: Done: Getting Results +2016-09-08 09:55:37,172 INFO: Classification on Fake database for View3 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588235294118 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588235294118 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0763602735229 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.51724137931 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.681818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.536561264822 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 +2016-09-08 09:55:37,173 INFO: Done: Result Analysis +2016-09-08 09:55:37,183 DEBUG: Done: Getting Results +2016-09-08 09:55:37,184 INFO: Classification on Fake database for View3 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 4, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.576923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.576923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0314347306731 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.681818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.514822134387 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 09:55:37,184 INFO: Done: Result Analysis +2016-09-08 09:55:37,335 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:37,335 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:37,335 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 09:55:37,335 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 09:55:37,335 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:37,335 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:37,336 DEBUG: Info: Shape X_train:(210, 6), Length of y_train:210 +2016-09-08 09:55:37,336 DEBUG: Info: Shape X_train:(210, 6), Length of y_train:210 +2016-09-08 09:55:37,336 DEBUG: Info: Shape X_test:(90, 6), Length of y_test:90 +2016-09-08 09:55:37,336 DEBUG: Info: Shape X_test:(90, 6), Length of y_test:90 +2016-09-08 09:55:37,336 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:37,336 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:37,336 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:37,336 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:37,365 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:37,365 DEBUG: Start: Training +2016-09-08 09:55:37,365 DEBUG: Info: Time for Training: 0.0312879085541[s] +2016-09-08 09:55:37,365 DEBUG: Done: Training +2016-09-08 09:55:37,366 DEBUG: Start: Predicting +2016-09-08 09:55:37,370 DEBUG: Done: Predicting +2016-09-08 09:55:37,371 DEBUG: Start: Getting Results +2016-09-08 09:55:37,421 DEBUG: Done: Getting Results +2016-09-08 09:55:37,422 INFO: Classification on Fake database for View3 with KNN + +accuracy_score on train : 0.590476190476 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.681481481481 + - Score on test : 0.601941747573 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.681481481481 + - Score on test : 0.601941747573 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.127350050081 + - Score on test : 0.100829966549 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.621621621622 + - Score on test : 0.525423728814 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.754098360656 + - Score on test : 0.704545454545 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.558867362146 + - Score on test : 0.547924901186 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 +2016-09-08 09:55:37,422 INFO: Done: Result Analysis +2016-09-08 09:55:37,439 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:37,439 DEBUG: Start: Training +2016-09-08 09:55:37,449 DEBUG: Info: Time for Training: 0.114704847336[s] +2016-09-08 09:55:37,449 DEBUG: Done: Training +2016-09-08 09:55:37,449 DEBUG: Start: Predicting +2016-09-08 09:55:37,452 DEBUG: Done: Predicting +2016-09-08 09:55:37,453 DEBUG: Start: Getting Results +2016-09-08 09:55:37,485 DEBUG: Done: Getting Results +2016-09-08 09:55:37,485 INFO: Classification on Fake database for View3 with RandomForest + +accuracy_score on train : 0.9 +accuracy_score on test : 0.577777777778 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 4, max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.9 + - Score on test : 0.577777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.910638297872 + - Score on test : 0.586956521739 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.910638297872 + - Score on test : 0.586956521739 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.1 + - Score on test : 0.422222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.9 + - Score on test : 0.577777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.800522373751 + - Score on test : 0.157426051223 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.946902654867 + - Score on test : 0.5625 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.877049180328 + - Score on test : 0.613636363636 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.904433681073 + - Score on test : 0.578557312253 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.1 + - Score on test : 0.422222222222 + + + Classification took 0:00:00 +2016-09-08 09:55:37,486 INFO: Done: Result Analysis +2016-09-08 09:55:37,579 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:37,579 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 09:55:37,579 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:55:37,579 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:37,579 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 09:55:37,579 DEBUG: Start: Determine Train/Test split +2016-09-08 09:55:37,579 DEBUG: Info: Shape X_train:(210, 6), Length of y_train:210 +2016-09-08 09:55:37,580 DEBUG: Info: Shape X_test:(90, 6), Length of y_test:90 +2016-09-08 09:55:37,580 DEBUG: Info: Shape X_train:(210, 6), Length of y_train:210 +2016-09-08 09:55:37,580 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:37,580 DEBUG: Info: Shape X_test:(90, 6), Length of y_test:90 +2016-09-08 09:55:37,580 DEBUG: Done: Determine Train/Test split +2016-09-08 09:55:37,580 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:37,580 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:55:37,625 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:37,625 DEBUG: Start: Training +2016-09-08 09:55:37,626 DEBUG: Info: Time for Training: 0.0474660396576[s] +2016-09-08 09:55:37,626 DEBUG: Done: Training +2016-09-08 09:55:37,626 DEBUG: Start: Predicting +2016-09-08 09:55:37,627 DEBUG: Done: RandomSearch best settings +2016-09-08 09:55:37,627 DEBUG: Start: Training +2016-09-08 09:55:37,643 DEBUG: Info: Time for Training: 0.0650768280029[s] +2016-09-08 09:55:37,644 DEBUG: Done: Training +2016-09-08 09:55:37,644 DEBUG: Start: Predicting +2016-09-08 09:55:37,647 DEBUG: Done: Predicting +2016-09-08 09:55:37,647 DEBUG: Start: Getting Results +2016-09-08 09:55:37,652 DEBUG: Done: Predicting +2016-09-08 09:55:37,652 DEBUG: Start: Getting Results +2016-09-08 09:55:37,675 DEBUG: Done: Getting Results +2016-09-08 09:55:37,675 INFO: Classification on Fake database for View3 with SGD + +accuracy_score on train : 0.590476190476 +accuracy_score on test : 0.488888888889 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : elasticnet + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.488888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.739393939394 + - Score on test : 0.65671641791 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.739393939394 + - Score on test : 0.65671641791 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.511111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.488888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.115457436228 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.586538461538 + - Score on test : 0.488888888889 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.511363636364 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.511111111111 + + + Classification took 0:00:00 +2016-09-08 09:55:37,675 INFO: Done: Result Analysis +2016-09-08 09:55:37,686 DEBUG: Done: Getting Results +2016-09-08 09:55:37,686 INFO: Classification on Fake database for View3 with SVMLinear + +accuracy_score on train : 0.495238095238 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.495238095238 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.595419847328 + - Score on test : 0.550458715596 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.595419847328 + - Score on test : 0.550458715596 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.504761904762 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.495238095238 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0682438863041 + - Score on test : -0.0882242643891 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.557142857143 + - Score on test : 0.461538461538 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.639344262295 + - Score on test : 0.681818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.467399403875 + - Score on test : 0.4604743083 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.504761904762 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:55:37,686 INFO: Done: Result Analysis +2016-09-08 09:55:37,978 INFO: ### Main Programm for Multiview Classification +2016-09-08 09:55:37,979 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Mumbo ; Cores : 1 +2016-09-08 09:55:37,980 INFO: Info: Shape of View0 :(300, 8) +2016-09-08 09:55:37,981 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 09:55:37,982 INFO: Info: Shape of View2 :(300, 7) +2016-09-08 09:55:37,983 INFO: Info: Shape of View3 :(300, 6) +2016-09-08 09:55:37,983 INFO: Done: Read Database Files +2016-09-08 09:55:37,983 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 09:55:37,987 INFO: ### Main Programm for Multiview Classification +2016-09-08 09:55:37,987 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 09:55:37,987 INFO: Done: Determine validation split +2016-09-08 09:55:37,987 INFO: Start: Determine 5 folds +2016-09-08 09:55:37,988 INFO: Info: Shape of View0 :(300, 8) +2016-09-08 09:55:37,988 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 09:55:37,989 INFO: Info: Shape of View2 :(300, 7) +2016-09-08 09:55:37,989 INFO: Info: Shape of View3 :(300, 6) +2016-09-08 09:55:37,989 INFO: Done: Read Database Files +2016-09-08 09:55:37,989 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 09:55:37,993 INFO: Done: Determine validation split +2016-09-08 09:55:37,993 INFO: Start: Determine 5 folds +2016-09-08 09:55:37,995 INFO: Info: Length of Learning Sets: 170 +2016-09-08 09:55:37,995 INFO: Info: Length of Testing Sets: 41 +2016-09-08 09:55:37,995 INFO: Info: Length of Validation Set: 89 +2016-09-08 09:55:37,995 INFO: Done: Determine folds +2016-09-08 09:55:37,995 INFO: Start: Learning with Mumbo and 5 folds +2016-09-08 09:55:37,996 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 09:55:37,996 DEBUG: Start: Random search for DecisionTree on View0 +2016-09-08 09:55:37,999 INFO: Info: Length of Learning Sets: 170 +2016-09-08 09:55:38,000 INFO: Info: Length of Testing Sets: 41 +2016-09-08 09:55:38,000 INFO: Info: Length of Validation Set: 89 +2016-09-08 09:55:38,000 INFO: Done: Determine folds +2016-09-08 09:55:38,000 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 09:55:38,000 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 09:55:38,000 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:55:38,054 DEBUG: Done: Random search for SGD +2016-09-08 09:55:38,054 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:55:38,107 DEBUG: Done: Random search for SGD +2016-09-08 09:55:38,107 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:55:38,157 DEBUG: Done: Random search for SGD +2016-09-08 09:55:38,157 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:55:38,212 DEBUG: Done: Random search for SGD +2016-09-08 09:55:38,284 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 09:55:38,284 INFO: Start: Classification +2016-09-08 09:55:38,284 INFO: Start: Fold number 1 +2016-09-08 09:55:38,311 INFO: Start: Classification +2016-09-08 09:55:38,336 INFO: Done: Fold number 1 +2016-09-08 09:55:38,337 INFO: Start: Fold number 2 +2016-09-08 09:55:38,363 INFO: Start: Classification +2016-09-08 09:55:38,389 INFO: Done: Fold number 2 +2016-09-08 09:55:38,389 INFO: Start: Fold number 3 +2016-09-08 09:55:38,416 INFO: Start: Classification +2016-09-08 09:55:38,443 INFO: Done: Fold number 3 +2016-09-08 09:55:38,443 INFO: Start: Fold number 4 +2016-09-08 09:55:38,470 INFO: Start: Classification +2016-09-08 09:55:38,496 INFO: Done: Fold number 4 +2016-09-08 09:55:38,496 INFO: Start: Fold number 5 +2016-09-08 09:55:38,523 INFO: Start: Classification +2016-09-08 09:55:38,549 INFO: Done: Fold number 5 +2016-09-08 09:55:38,549 INFO: Done: Classification +2016-09-08 09:55:38,549 INFO: Info: Time for Classification: 0[s] +2016-09-08 09:55:38,550 INFO: Start: Result Analysis for Fusion +2016-09-08 09:55:38,681 INFO: Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 57.1764705882 + -On Test : 52.6829268293 + -On Validation : 52.1348314607 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Bayesian Inference using a weight for each view : 0.322967175445, 0.0184701333132, 0.322597810111, 0.335964881131 + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : modified_huber, penalty : elasticnet + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 09:55:38,682 INFO: Done: Result Analysis diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..0d0cf3ff0c83bca73e37fc715e290390079bbb69 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View0 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 4, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.505263157895 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.505263157895 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0418655345164 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.470588235294 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.545454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.479249011858 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..9f126a9ea93eea83d8098254d24706eaafb2b497 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.484210526316 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.484210526316 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0867214643554 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.450980392157 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522727272727 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.457015810277 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..79be41beb0284d7bf6eb3848621838655d406c4c --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with KNN + +accuracy_score on train : 0.661904761905 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.661904761905 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.727969348659 + - Score on test : 0.534653465347 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.727969348659 + - Score on test : 0.534653465347 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.338095238095 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.661904761905 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.290672377783 + - Score on test : -0.0399755963154 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.68345323741 + - Score on test : 0.473684210526 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.77868852459 + - Score on test : 0.613636363636 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.639344262295 + - Score on test : 0.480731225296 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.338095238095 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..bbd624877d4bfc10a8b89482d2ec1a7829577c75 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with RandomForest + +accuracy_score on train : 0.895238095238 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 4, max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.895238095238 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.905982905983 + - Score on test : 0.515463917526 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.905982905983 + - Score on test : 0.515463917526 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.104761904762 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.895238095238 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.791868570857 + - Score on test : -0.0411594726194 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.946428571429 + - Score on test : 0.471698113208 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.868852459016 + - Score on test : 0.568181818182 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.900335320417 + - Score on test : 0.479743083004 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.104761904762 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..fd413a5d803bf18dd794b651f84d3eeb305d6c94 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SGD + +accuracy_score on train : 0.614285714286 +accuracy_score on test : 0.466666666667 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : elasticnet + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.614285714286 + - Score on test : 0.466666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.749226006192 + - Score on test : 0.630769230769 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.749226006192 + - Score on test : 0.630769230769 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.385714285714 + - Score on test : 0.533333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.614285714286 + - Score on test : 0.466666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.201498784613 + - Score on test : -0.112653159931 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.601990049751 + - Score on test : 0.476744186047 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.991803278689 + - Score on test : 0.931818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.541356184799 + - Score on test : 0.476778656126 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.385714285714 + - Score on test : 0.533333333333 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..7b2f009cb9959d5bd993cf1b27c37272d3ffaf39 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMLinear + +accuracy_score on train : 0.490476190476 +accuracy_score on test : 0.377777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.490476190476 + - Score on test : 0.377777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.563265306122 + - Score on test : 0.416666666667 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.563265306122 + - Score on test : 0.416666666667 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.509523809524 + - Score on test : 0.622222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.490476190476 + - Score on test : 0.377777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0481411286791 + - Score on test : -0.244017569898 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.560975609756 + - Score on test : 0.384615384615 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.565573770492 + - Score on test : 0.454545454545 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.475968703428 + - Score on test : 0.379446640316 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.509523809524 + - Score on test : 0.622222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..db536ff68f9249c6cab5a9be1ebf6c21dbb9de58 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.468085106383 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.468085106383 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.109345881217 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.44 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.445652173913 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..4251899bb7a8bf556dee03c2c785c4ac0e48d076 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095534Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 8) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.616822429907 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.616822429907 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.106710653456 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.52380952381 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.75 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.548913043478 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..a32b018a2505e28f037bf72ff427fc03005ce552 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View1 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 4, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.48 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.48 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.154858431981 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.428571428571 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.545454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.424901185771 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..b096d574bd224677355e3b214b9cfe60865aabed --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.48 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.48 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.154858431981 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.428571428571 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.545454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.424901185771 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..376343e68cd0ba94ded748ff13a5313cc2f34cb9 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with KNN + +accuracy_score on train : 0.571428571429 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.6484375 + - Score on test : 0.541666666667 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.6484375 + - Score on test : 0.541666666667 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.428571428571 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.103477711187 + - Score on test : 0.0260018722022 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.619402985075 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.680327868852 + - Score on test : 0.590909090909 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.550391207154 + - Score on test : 0.512845849802 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.428571428571 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..c3e2e2995350f4e7c2570de375e0fe6d36fb5526 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with RandomForest + +accuracy_score on train : 0.9 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 4, max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.9 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.909090909091 + - Score on test : 0.356164383562 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.909090909091 + - Score on test : 0.356164383562 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.1 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.9 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.805030105216 + - Score on test : -0.0560191732057 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.963302752294 + - Score on test : 0.448275862069 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.860655737705 + - Score on test : 0.295454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.907600596125 + - Score on test : 0.473814229249 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.1 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..1bd711b474351f5d09ddf4a343c4f7ee989d2036 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SGD + +accuracy_score on train : 0.609523809524 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : elasticnet + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.609523809524 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.745341614907 + - Score on test : 0.661417322835 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.745341614907 + - Score on test : 0.661417322835 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.390476190476 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.609523809524 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.172644893682 + - Score on test : 0.118036588599 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.6 + - Score on test : 0.506024096386 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.983606557377 + - Score on test : 0.954545454545 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.537257824143 + - Score on test : 0.53162055336 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.390476190476 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..6b6c9c97508ff479293091869f19742e26369f0a --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMLinear + +accuracy_score on train : 0.542857142857 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.542857142857 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.606557377049 + - Score on test : 0.584905660377 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.606557377049 + - Score on test : 0.584905660377 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.457142857143 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.542857142857 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0611028315946 + - Score on test : 0.0330758927464 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.606557377049 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.606557377049 + - Score on test : 0.704545454545 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.530551415797 + - Score on test : 0.515316205534 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.457142857143 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..5bc5811d3a8eecb74b0763c70685c8d806b52e8b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.358208955224 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.358208955224 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0385036888617 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.521739130435 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.272727272727 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.516798418972 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..ccd0f1ad914c25ad47b37f5f969f4202f2a8de66 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095535Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.645161290323 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.645161290323 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0628694613462 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.909090909091 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.51976284585 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..9deb12f3b14d4f8565c6168fa4a139bddf91fb16 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View2 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 4, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.436781609195 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.436781609195 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0899876638096 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.441860465116 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.431818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455039525692 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..183d09191497b8ab3804d324b510249818998073 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.466666666667 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.441860465116 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.441860465116 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.533333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.068316965625 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.452380952381 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.431818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.465909090909 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.533333333333 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..8cf553cc1128b3958539e214301a382219781461 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with KNN + +accuracy_score on train : 0.619047619048 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.619047619048 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.705882352941 + - Score on test : 0.619469026549 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.705882352941 + - Score on test : 0.619469026549 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.380952380952 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.619047619048 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.189221481343 + - Score on test : 0.0665679839847 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.64 + - Score on test : 0.507246376812 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.786885245902 + - Score on test : 0.795454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.586624441133 + - Score on test : 0.528162055336 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.380952380952 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..f74150b10001487507af804c177f45b9af53b193 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with RandomForest + +accuracy_score on train : 0.890476190476 +accuracy_score on test : 0.4 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 4, max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.890476190476 + - Score on test : 0.4 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.904564315353 + - Score on test : 0.357142857143 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.904564315353 + - Score on test : 0.357142857143 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.109523809524 + - Score on test : 0.6 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.890476190476 + - Score on test : 0.4 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.77645053118 + - Score on test : -0.20378096045 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.915966386555 + - Score on test : 0.375 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.893442622951 + - Score on test : 0.340909090909 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.889903129657 + - Score on test : 0.39871541502 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.109523809524 + - Score on test : 0.6 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..608729cd17cff989b5994f699bcc2c9f7d249468 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SGD + +accuracy_score on train : 0.590476190476 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : elasticnet + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.715231788079 + - Score on test : 0.676923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.715231788079 + - Score on test : 0.676923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0945615027077 + - Score on test : 0.210925065403 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.6 + - Score on test : 0.511627906977 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.885245901639 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.533532041729 + - Score on test : 0.54347826087 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..59671aeef690b38b65227b5dd243edfa5b9b1e5b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMLinear + +accuracy_score on train : 0.533333333333 +accuracy_score on test : 0.433333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.433333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.604838709677 + - Score on test : 0.484848484848 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.604838709677 + - Score on test : 0.484848484848 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.566666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.433333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0354605635154 + - Score on test : -0.131720304791 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.595238095238 + - Score on test : 0.436363636364 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.614754098361 + - Score on test : 0.545454545455 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.517604321908 + - Score on test : 0.435770750988 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.566666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..a085f7609062a0cf55e082cfc7420a115dad6758 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.4 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.357142857143 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.357142857143 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.6 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.20378096045 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.375 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.340909090909 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.39871541502 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.6 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..cf24b718b7d42ee523ad9050390a3c2e19696090 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095536Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.466666666667 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 7) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.538461538462 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.538461538462 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.533333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0628694613462 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.636363636364 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.470355731225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.533333333333 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..b93b2e14fa71d3dfb0ae9bead5ea8c59f93ea5b9 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View3 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 4, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.576923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.576923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0314347306731 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.681818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.514822134387 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..ed86c3be7a62566d0a9aeb7b157f834ad4429c2d --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588235294118 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588235294118 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0763602735229 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.51724137931 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.681818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.536561264822 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..81a8c71b0206a996bae52dbc3db9c74f8f8ceb77 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with KNN + +accuracy_score on train : 0.590476190476 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.681481481481 + - Score on test : 0.601941747573 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.681481481481 + - Score on test : 0.601941747573 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.127350050081 + - Score on test : 0.100829966549 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.621621621622 + - Score on test : 0.525423728814 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.754098360656 + - Score on test : 0.704545454545 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.558867362146 + - Score on test : 0.547924901186 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..5a7ef634e77ff082d71a4580ab038c25426e46c2 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with RandomForest + +accuracy_score on train : 0.9 +accuracy_score on test : 0.577777777778 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 4, max_depth : 20 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.9 + - Score on test : 0.577777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.910638297872 + - Score on test : 0.586956521739 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.910638297872 + - Score on test : 0.586956521739 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.1 + - Score on test : 0.422222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.9 + - Score on test : 0.577777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.800522373751 + - Score on test : 0.157426051223 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.946902654867 + - Score on test : 0.5625 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.877049180328 + - Score on test : 0.613636363636 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.904433681073 + - Score on test : 0.578557312253 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.1 + - Score on test : 0.422222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..616ff06f5c3552206bb0e28675b677ea6d137aac --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with SGD + +accuracy_score on train : 0.590476190476 +accuracy_score on test : 0.488888888889 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : elasticnet + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.488888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.739393939394 + - Score on test : 0.65671641791 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.739393939394 + - Score on test : 0.65671641791 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.511111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.590476190476 + - Score on test : 0.488888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.115457436228 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.586538461538 + - Score on test : 0.488888888889 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.511363636364 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.409523809524 + - Score on test : 0.511111111111 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..226afbe88c8d26e49858a82280f274251102f2e3 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095537Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with SVMLinear + +accuracy_score on train : 0.495238095238 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 6) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 6107 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.495238095238 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.595419847328 + - Score on test : 0.550458715596 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.595419847328 + - Score on test : 0.550458715596 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.504761904762 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.495238095238 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0682438863041 + - Score on test : -0.0882242643891 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.557142857143 + - Score on test : 0.461538461538 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.639344262295 + - Score on test : 0.681818181818 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.467399403875 + - Score on test : 0.4604743083 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.504761904762 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095538Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095538Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d2af511daa99659d08543c83537f0fa618a2ebef --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095538Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,32 @@ + Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 57.1764705882 + -On Test : 52.6829268293 + -On Validation : 52.1348314607 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Bayesian Inference using a weight for each view : 0.322967175445, 0.0184701333132, 0.322597810111, 0.335964881131 + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : modified_huber, penalty : elasticnet + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095622-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log b/Code/MonoMutliViewClassifiers/Results/20160908-095622-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log new file mode 100644 index 0000000000000000000000000000000000000000..e8dd4634725d66617aab8ac7917d7f46df47a448 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095622-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log @@ -0,0 +1,2394 @@ +2016-09-08 09:56:22,282 DEBUG: Start: Creating 2 temporary datasets for multiprocessing +2016-09-08 09:56:22,282 WARNING: WARNING : /!\ This may use a lot of HDD storage space : 0.0001661875 Gbytes /!\ +2016-09-08 09:56:27,294 DEBUG: Start: Creating datasets for multiprocessing +2016-09-08 09:56:27,298 INFO: Start: Finding all available mono- & multiview algorithms +2016-09-08 09:56:27,351 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:27,351 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:56:27,351 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:27,352 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:27,352 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:56:27,352 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:27,352 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:27,352 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:27,352 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:27,353 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:27,353 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:27,353 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:27,353 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:27,354 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:27,389 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:27,389 DEBUG: Start: Training +2016-09-08 09:56:27,391 DEBUG: Info: Time for Training: 0.0395510196686[s] +2016-09-08 09:56:27,391 DEBUG: Done: Training +2016-09-08 09:56:27,391 DEBUG: Start: Predicting +2016-09-08 09:56:27,393 DEBUG: Done: Predicting +2016-09-08 09:56:27,393 DEBUG: Start: Getting Results +2016-09-08 09:56:27,403 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:27,403 DEBUG: Start: Training +2016-09-08 09:56:27,407 DEBUG: Info: Time for Training: 0.0563409328461[s] +2016-09-08 09:56:27,407 DEBUG: Done: Training +2016-09-08 09:56:27,407 DEBUG: Start: Predicting +2016-09-08 09:56:27,410 DEBUG: Done: Predicting +2016-09-08 09:56:27,410 DEBUG: Start: Getting Results +2016-09-08 09:56:27,440 DEBUG: Done: Getting Results +2016-09-08 09:56:27,440 INFO: Classification on Fake database for View0 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.377777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.377777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.348837209302 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.348837209302 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.622222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.377777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.245415911539 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.365853658537 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.376804380289 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.622222222222 + + + Classification took 0:00:00 +2016-09-08 09:56:27,441 INFO: Done: Result Analysis +2016-09-08 09:56:27,452 DEBUG: Done: Getting Results +2016-09-08 09:56:27,452 INFO: Classification on Fake database for View0 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.377777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 10, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.377777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.622222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.377777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.249628898234 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.325581395349 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.341463414634 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.37481333997 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.622222222222 + + + Classification took 0:00:00 +2016-09-08 09:56:27,452 INFO: Done: Result Analysis +2016-09-08 09:56:27,595 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:27,595 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:27,595 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 09:56:27,595 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 09:56:27,595 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:27,595 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:27,596 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:27,596 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:27,596 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:27,596 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:27,596 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:27,596 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:27,596 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:27,596 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:27,627 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:27,627 DEBUG: Start: Training +2016-09-08 09:56:27,628 DEBUG: Info: Time for Training: 0.0337619781494[s] +2016-09-08 09:56:27,628 DEBUG: Done: Training +2016-09-08 09:56:27,628 DEBUG: Start: Predicting +2016-09-08 09:56:27,635 DEBUG: Done: Predicting +2016-09-08 09:56:27,635 DEBUG: Start: Getting Results +2016-09-08 09:56:27,676 DEBUG: Done: Getting Results +2016-09-08 09:56:27,676 INFO: Classification on Fake database for View0 with KNN + +accuracy_score on train : 0.609523809524 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 42 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.609523809524 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.467532467532 + - Score on test : 0.212121212121 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.467532467532 + - Score on test : 0.212121212121 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.390476190476 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.609523809524 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.237135686833 + - Score on test : -0.218615245335 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.692307692308 + - Score on test : 0.28 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.352941176471 + - Score on test : 0.170731707317 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.602396514161 + - Score on test : 0.401692384271 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.390476190476 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 +2016-09-08 09:56:27,676 INFO: Done: Result Analysis +2016-09-08 09:56:27,921 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:27,921 DEBUG: Start: Training +2016-09-08 09:56:27,970 DEBUG: Info: Time for Training: 0.375941991806[s] +2016-09-08 09:56:27,970 DEBUG: Done: Training +2016-09-08 09:56:27,970 DEBUG: Start: Predicting +2016-09-08 09:56:27,976 DEBUG: Done: Predicting +2016-09-08 09:56:27,977 DEBUG: Start: Getting Results +2016-09-08 09:56:28,010 DEBUG: Done: Getting Results +2016-09-08 09:56:28,011 INFO: Classification on Fake database for View0 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 19, max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.423529411765 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.423529411765 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0912478416452 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.409090909091 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.439024390244 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.454206072673 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:56:28,011 INFO: Done: Result Analysis +2016-09-08 09:56:28,143 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:28,143 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 09:56:28,143 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:28,143 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:28,143 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 09:56:28,144 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:28,144 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:28,144 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:28,144 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:28,144 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:28,144 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:28,144 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:28,144 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:28,145 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:28,189 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:28,189 DEBUG: Start: Training +2016-09-08 09:56:28,190 DEBUG: Info: Time for Training: 0.0481970310211[s] +2016-09-08 09:56:28,190 DEBUG: Done: Training +2016-09-08 09:56:28,190 DEBUG: Start: Predicting +2016-09-08 09:56:28,195 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:28,195 DEBUG: Start: Training +2016-09-08 09:56:28,214 DEBUG: Info: Time for Training: 0.0711450576782[s] +2016-09-08 09:56:28,214 DEBUG: Done: Training +2016-09-08 09:56:28,214 DEBUG: Start: Predicting +2016-09-08 09:56:28,216 DEBUG: Done: Predicting +2016-09-08 09:56:28,216 DEBUG: Start: Getting Results +2016-09-08 09:56:28,217 DEBUG: Done: Predicting +2016-09-08 09:56:28,217 DEBUG: Start: Getting Results +2016-09-08 09:56:28,239 DEBUG: Done: Getting Results +2016-09-08 09:56:28,239 INFO: Classification on Fake database for View0 with SGD + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : l1 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:56:28,239 INFO: Done: Result Analysis +2016-09-08 09:56:28,250 DEBUG: Done: Getting Results +2016-09-08 09:56:28,250 INFO: Classification on Fake database for View0 with SVMLinear + +accuracy_score on train : 0.52380952381 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.52380952381 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.532710280374 + - Score on test : 0.459770114943 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.532710280374 + - Score on test : 0.459770114943 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.47619047619 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.52380952381 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0496545019224 + - Score on test : -0.0426484477255 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.508928571429 + - Score on test : 0.434782608696 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.558823529412 + - Score on test : 0.487804878049 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.524782135076 + - Score on test : 0.478596316575 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.47619047619 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 +2016-09-08 09:56:28,250 INFO: Done: Result Analysis +2016-09-08 09:56:28,387 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:28,387 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:28,387 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 09:56:28,387 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 09:56:28,387 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:28,387 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:28,388 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:28,388 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:28,388 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:28,388 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:28,388 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:28,388 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:28,388 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:28,388 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:28,433 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:28,433 DEBUG: Start: Training +2016-09-08 09:56:28,438 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:28,438 DEBUG: Start: Training +2016-09-08 09:56:28,450 DEBUG: Info: Time for Training: 0.0634729862213[s] +2016-09-08 09:56:28,450 DEBUG: Done: Training +2016-09-08 09:56:28,450 DEBUG: Start: Predicting +2016-09-08 09:56:28,455 DEBUG: Info: Time for Training: 0.0687439441681[s] +2016-09-08 09:56:28,455 DEBUG: Done: Training +2016-09-08 09:56:28,455 DEBUG: Start: Predicting +2016-09-08 09:56:28,455 DEBUG: Done: Predicting +2016-09-08 09:56:28,455 DEBUG: Start: Getting Results +2016-09-08 09:56:28,459 DEBUG: Done: Predicting +2016-09-08 09:56:28,459 DEBUG: Start: Getting Results +2016-09-08 09:56:28,491 DEBUG: Done: Getting Results +2016-09-08 09:56:28,491 INFO: Classification on Fake database for View0 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.409090909091 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.409090909091 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.152357995542 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.382978723404 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.439024390244 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.423593827775 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 +2016-09-08 09:56:28,492 INFO: Done: Result Analysis +2016-09-08 09:56:28,505 DEBUG: Done: Getting Results +2016-09-08 09:56:28,505 INFO: Classification on Fake database for View0 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.475 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.475 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0555284586866 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.487179487179 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.463414634146 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.52762568442 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 +2016-09-08 09:56:28,505 INFO: Done: Result Analysis +2016-09-08 09:56:28,637 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:28,638 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:56:28,638 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:28,638 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:28,638 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:56:28,638 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:28,638 DEBUG: Info: Shape X_train:(210, 19), Length of y_train:210 +2016-09-08 09:56:28,638 DEBUG: Info: Shape X_test:(90, 19), Length of y_test:90 +2016-09-08 09:56:28,638 DEBUG: Info: Shape X_train:(210, 19), Length of y_train:210 +2016-09-08 09:56:28,639 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:28,639 DEBUG: Info: Shape X_test:(90, 19), Length of y_test:90 +2016-09-08 09:56:28,639 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:28,639 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:28,639 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:28,677 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:28,677 DEBUG: Start: Training +2016-09-08 09:56:28,680 DEBUG: Info: Time for Training: 0.0427668094635[s] +2016-09-08 09:56:28,680 DEBUG: Done: Training +2016-09-08 09:56:28,680 DEBUG: Start: Predicting +2016-09-08 09:56:28,683 DEBUG: Done: Predicting +2016-09-08 09:56:28,683 DEBUG: Start: Getting Results +2016-09-08 09:56:28,691 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:28,691 DEBUG: Start: Training +2016-09-08 09:56:28,696 DEBUG: Info: Time for Training: 0.0597720146179[s] +2016-09-08 09:56:28,696 DEBUG: Done: Training +2016-09-08 09:56:28,696 DEBUG: Start: Predicting +2016-09-08 09:56:28,699 DEBUG: Done: Predicting +2016-09-08 09:56:28,699 DEBUG: Start: Getting Results +2016-09-08 09:56:28,734 DEBUG: Done: Getting Results +2016-09-08 09:56:28,734 INFO: Classification on Fake database for View1 with DecisionTree + +accuracy_score on train : 0.957142857143 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.957142857143 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.957746478873 + - Score on test : 0.444444444444 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.957746478873 + - Score on test : 0.444444444444 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0428571428571 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.957142857143 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.917792101918 + - Score on test : -0.104031856645 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.918918918919 + - Score on test : 0.408163265306 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.487804878049 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.958333333333 + - Score on test : 0.447984071677 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0428571428571 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 +2016-09-08 09:56:28,734 INFO: Done: Result Analysis +2016-09-08 09:56:28,741 DEBUG: Done: Getting Results +2016-09-08 09:56:28,741 INFO: Classification on Fake database for View1 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 10, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.119960179194 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.440019910403 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 +2016-09-08 09:56:28,741 INFO: Done: Result Analysis +2016-09-08 09:56:28,882 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:28,883 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 09:56:28,883 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:28,883 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:28,883 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 09:56:28,884 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:28,884 DEBUG: Info: Shape X_train:(210, 19), Length of y_train:210 +2016-09-08 09:56:28,884 DEBUG: Info: Shape X_test:(90, 19), Length of y_test:90 +2016-09-08 09:56:28,884 DEBUG: Info: Shape X_train:(210, 19), Length of y_train:210 +2016-09-08 09:56:28,884 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:28,885 DEBUG: Info: Shape X_test:(90, 19), Length of y_test:90 +2016-09-08 09:56:28,885 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:28,885 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:28,885 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:28,919 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:28,919 DEBUG: Start: Training +2016-09-08 09:56:28,920 DEBUG: Info: Time for Training: 0.0378148555756[s] +2016-09-08 09:56:28,920 DEBUG: Done: Training +2016-09-08 09:56:28,920 DEBUG: Start: Predicting +2016-09-08 09:56:28,927 DEBUG: Done: Predicting +2016-09-08 09:56:28,927 DEBUG: Start: Getting Results +2016-09-08 09:56:28,967 DEBUG: Done: Getting Results +2016-09-08 09:56:28,967 INFO: Classification on Fake database for View1 with KNN + +accuracy_score on train : 0.561904761905 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 42 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.561904761905 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.6 + - Score on test : 0.559139784946 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.6 + - Score on test : 0.559139784946 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.438095238095 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.561904761905 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.133359904768 + - Score on test : 0.104395047556 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.5390625 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.676470588235 + - Score on test : 0.634146341463 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.565087145969 + - Score on test : 0.551767048283 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.438095238095 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 +2016-09-08 09:56:28,967 INFO: Done: Result Analysis +2016-09-08 09:56:29,211 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:29,212 DEBUG: Start: Training +2016-09-08 09:56:29,260 DEBUG: Info: Time for Training: 0.377947092056[s] +2016-09-08 09:56:29,261 DEBUG: Done: Training +2016-09-08 09:56:29,261 DEBUG: Start: Predicting +2016-09-08 09:56:29,267 DEBUG: Done: Predicting +2016-09-08 09:56:29,267 DEBUG: Start: Getting Results +2016-09-08 09:56:29,300 DEBUG: Done: Getting Results +2016-09-08 09:56:29,300 INFO: Classification on Fake database for View1 with RandomForest + +accuracy_score on train : 0.995238095238 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 19, max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.995073891626 + - Score on test : 0.516853932584 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.995073891626 + - Score on test : 0.516853932584 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.990510833227 + - Score on test : 0.0506833064614 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.479166666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.990196078431 + - Score on test : 0.560975609756 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.995098039216 + - Score on test : 0.525385764062 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 +2016-09-08 09:56:29,300 INFO: Done: Result Analysis +2016-09-08 09:56:29,434 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:29,435 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 09:56:29,435 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:29,435 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:29,435 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 09:56:29,435 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:29,436 DEBUG: Info: Shape X_train:(210, 19), Length of y_train:210 +2016-09-08 09:56:29,436 DEBUG: Info: Shape X_test:(90, 19), Length of y_test:90 +2016-09-08 09:56:29,436 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:29,436 DEBUG: Info: Shape X_train:(210, 19), Length of y_train:210 +2016-09-08 09:56:29,436 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:29,436 DEBUG: Info: Shape X_test:(90, 19), Length of y_test:90 +2016-09-08 09:56:29,437 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:29,437 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:29,483 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:29,483 DEBUG: Start: Training +2016-09-08 09:56:29,484 DEBUG: Info: Time for Training: 0.0506761074066[s] +2016-09-08 09:56:29,484 DEBUG: Done: Training +2016-09-08 09:56:29,484 DEBUG: Start: Predicting +2016-09-08 09:56:29,491 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:29,491 DEBUG: Start: Training +2016-09-08 09:56:29,511 DEBUG: Done: Predicting +2016-09-08 09:56:29,511 DEBUG: Start: Getting Results +2016-09-08 09:56:29,520 DEBUG: Info: Time for Training: 0.0860919952393[s] +2016-09-08 09:56:29,520 DEBUG: Done: Training +2016-09-08 09:56:29,521 DEBUG: Start: Predicting +2016-09-08 09:56:29,526 DEBUG: Done: Predicting +2016-09-08 09:56:29,526 DEBUG: Start: Getting Results +2016-09-08 09:56:29,539 DEBUG: Done: Getting Results +2016-09-08 09:56:29,539 INFO: Classification on Fake database for View1 with SGD + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : l1 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:56:29,540 INFO: Done: Result Analysis +2016-09-08 09:56:29,553 DEBUG: Done: Getting Results +2016-09-08 09:56:29,554 INFO: Classification on Fake database for View1 with SVMLinear + +accuracy_score on train : 0.514285714286 +accuracy_score on test : 0.577777777778 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.514285714286 + - Score on test : 0.577777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.514285714286 + - Score on test : 0.547619047619 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.514285714286 + - Score on test : 0.547619047619 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.485714285714 + - Score on test : 0.422222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.514285714286 + - Score on test : 0.577777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0294117647059 + - Score on test : 0.152357995542 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.53488372093 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.529411764706 + - Score on test : 0.560975609756 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.514705882353 + - Score on test : 0.576406172225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.485714285714 + - Score on test : 0.422222222222 + + + Classification took 0:00:00 +2016-09-08 09:56:29,554 INFO: Done: Result Analysis +2016-09-08 09:56:29,683 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:29,684 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:29,684 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 09:56:29,684 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 09:56:29,684 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:29,684 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:29,685 DEBUG: Info: Shape X_train:(210, 19), Length of y_train:210 +2016-09-08 09:56:29,685 DEBUG: Info: Shape X_train:(210, 19), Length of y_train:210 +2016-09-08 09:56:29,685 DEBUG: Info: Shape X_test:(90, 19), Length of y_test:90 +2016-09-08 09:56:29,685 DEBUG: Info: Shape X_test:(90, 19), Length of y_test:90 +2016-09-08 09:56:29,686 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:29,686 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:29,686 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:29,686 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:29,756 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:29,756 DEBUG: Start: Training +2016-09-08 09:56:29,762 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:29,762 DEBUG: Start: Training +2016-09-08 09:56:29,781 DEBUG: Info: Time for Training: 0.0979940891266[s] +2016-09-08 09:56:29,781 DEBUG: Done: Training +2016-09-08 09:56:29,781 DEBUG: Start: Predicting +2016-09-08 09:56:29,788 DEBUG: Info: Time for Training: 0.105060100555[s] +2016-09-08 09:56:29,788 DEBUG: Done: Training +2016-09-08 09:56:29,788 DEBUG: Start: Predicting +2016-09-08 09:56:29,790 DEBUG: Done: Predicting +2016-09-08 09:56:29,790 DEBUG: Start: Getting Results +2016-09-08 09:56:29,793 DEBUG: Done: Predicting +2016-09-08 09:56:29,794 DEBUG: Start: Getting Results +2016-09-08 09:56:29,827 DEBUG: Done: Getting Results +2016-09-08 09:56:29,827 DEBUG: Done: Getting Results +2016-09-08 09:56:29,827 INFO: Classification on Fake database for View1 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.394366197183 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.394366197183 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0157759322964 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.341463414634 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.507466401195 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 +2016-09-08 09:56:29,827 INFO: Classification on Fake database for View1 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0193709711057 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.44 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.268292682927 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.491289198606 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 09:56:29,827 INFO: Done: Result Analysis +2016-09-08 09:56:29,828 INFO: Done: Result Analysis +2016-09-08 09:56:29,934 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:29,934 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:56:29,934 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:29,934 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:29,935 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:56:29,935 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:29,935 DEBUG: Info: Shape X_train:(210, 20), Length of y_train:210 +2016-09-08 09:56:29,935 DEBUG: Info: Shape X_test:(90, 20), Length of y_test:90 +2016-09-08 09:56:29,936 DEBUG: Info: Shape X_train:(210, 20), Length of y_train:210 +2016-09-08 09:56:29,936 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:29,936 DEBUG: Info: Shape X_test:(90, 20), Length of y_test:90 +2016-09-08 09:56:29,936 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:29,936 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:29,936 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:29,974 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:29,974 DEBUG: Start: Training +2016-09-08 09:56:29,976 DEBUG: Info: Time for Training: 0.0422959327698[s] +2016-09-08 09:56:29,976 DEBUG: Done: Training +2016-09-08 09:56:29,976 DEBUG: Start: Predicting +2016-09-08 09:56:29,979 DEBUG: Done: Predicting +2016-09-08 09:56:29,979 DEBUG: Start: Getting Results +2016-09-08 09:56:29,988 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:29,988 DEBUG: Start: Training +2016-09-08 09:56:29,994 DEBUG: Info: Time for Training: 0.0606279373169[s] +2016-09-08 09:56:29,994 DEBUG: Done: Training +2016-09-08 09:56:29,994 DEBUG: Start: Predicting +2016-09-08 09:56:29,997 DEBUG: Done: Predicting +2016-09-08 09:56:29,997 DEBUG: Start: Getting Results +2016-09-08 09:56:30,029 DEBUG: Done: Getting Results +2016-09-08 09:56:30,029 INFO: Classification on Fake database for View2 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.433333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.385542168675 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.385542168675 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.140124435511 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.380952380952 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.429815828771 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + + + Classification took 0:00:00 +2016-09-08 09:56:30,029 INFO: Done: Result Analysis +2016-09-08 09:56:30,043 DEBUG: Done: Getting Results +2016-09-08 09:56:30,043 INFO: Classification on Fake database for View2 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 10, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.375 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.375 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.124563839757 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.384615384615 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.365853658537 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.438028870085 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 +2016-09-08 09:56:30,044 INFO: Done: Result Analysis +2016-09-08 09:56:30,185 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:30,186 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 09:56:30,186 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:30,186 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:30,186 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 09:56:30,186 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:30,187 DEBUG: Info: Shape X_train:(210, 20), Length of y_train:210 +2016-09-08 09:56:30,187 DEBUG: Info: Shape X_train:(210, 20), Length of y_train:210 +2016-09-08 09:56:30,187 DEBUG: Info: Shape X_test:(90, 20), Length of y_test:90 +2016-09-08 09:56:30,187 DEBUG: Info: Shape X_test:(90, 20), Length of y_test:90 +2016-09-08 09:56:30,187 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:30,187 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:30,187 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:30,188 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:30,236 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:30,237 DEBUG: Start: Training +2016-09-08 09:56:30,238 DEBUG: Info: Time for Training: 0.0538048744202[s] +2016-09-08 09:56:30,238 DEBUG: Done: Training +2016-09-08 09:56:30,238 DEBUG: Start: Predicting +2016-09-08 09:56:30,249 DEBUG: Done: Predicting +2016-09-08 09:56:30,249 DEBUG: Start: Getting Results +2016-09-08 09:56:30,288 DEBUG: Done: Getting Results +2016-09-08 09:56:30,288 INFO: Classification on Fake database for View2 with KNN + +accuracy_score on train : 0.542857142857 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 42 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.542857142857 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.323943661972 + - Score on test : 0.327272727273 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.323943661972 + - Score on test : 0.327272727273 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.457142857143 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.542857142857 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0866552427925 + - Score on test : 0.161417724438 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.575 + - Score on test : 0.642857142857 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.225490196078 + - Score on test : 0.219512195122 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.534041394336 + - Score on test : 0.558735689398 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.457142857143 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 +2016-09-08 09:56:30,289 INFO: Done: Result Analysis +2016-09-08 09:56:30,545 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:30,545 DEBUG: Start: Training +2016-09-08 09:56:30,594 DEBUG: Info: Time for Training: 0.40927195549[s] +2016-09-08 09:56:30,594 DEBUG: Done: Training +2016-09-08 09:56:30,594 DEBUG: Start: Predicting +2016-09-08 09:56:30,600 DEBUG: Done: Predicting +2016-09-08 09:56:30,600 DEBUG: Start: Getting Results +2016-09-08 09:56:30,633 DEBUG: Done: Getting Results +2016-09-08 09:56:30,633 INFO: Classification on Fake database for View2 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 19, max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.55421686747 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.55421686747 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.172919516163 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.547619047619 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.560975609756 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.586610253858 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 +2016-09-08 09:56:30,633 INFO: Done: Result Analysis +2016-09-08 09:56:30,727 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:30,727 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 09:56:30,727 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:30,727 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:30,728 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 09:56:30,728 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:30,728 DEBUG: Info: Shape X_train:(210, 20), Length of y_train:210 +2016-09-08 09:56:30,728 DEBUG: Info: Shape X_test:(90, 20), Length of y_test:90 +2016-09-08 09:56:30,728 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:30,728 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:30,728 DEBUG: Info: Shape X_train:(210, 20), Length of y_train:210 +2016-09-08 09:56:30,728 DEBUG: Info: Shape X_test:(90, 20), Length of y_test:90 +2016-09-08 09:56:30,729 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:30,729 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:30,773 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:30,773 DEBUG: Start: Training +2016-09-08 09:56:30,774 DEBUG: Info: Time for Training: 0.0476858615875[s] +2016-09-08 09:56:30,774 DEBUG: Done: Training +2016-09-08 09:56:30,774 DEBUG: Start: Predicting +2016-09-08 09:56:30,779 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:30,779 DEBUG: Start: Training +2016-09-08 09:56:30,796 DEBUG: Done: Predicting +2016-09-08 09:56:30,796 DEBUG: Start: Getting Results +2016-09-08 09:56:30,800 DEBUG: Info: Time for Training: 0.0737199783325[s] +2016-09-08 09:56:30,800 DEBUG: Done: Training +2016-09-08 09:56:30,801 DEBUG: Start: Predicting +2016-09-08 09:56:30,804 DEBUG: Done: Predicting +2016-09-08 09:56:30,804 DEBUG: Start: Getting Results +2016-09-08 09:56:30,819 DEBUG: Done: Getting Results +2016-09-08 09:56:30,819 INFO: Classification on Fake database for View2 with SGD + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : l1 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:56:30,819 INFO: Done: Result Analysis +2016-09-08 09:56:30,833 DEBUG: Done: Getting Results +2016-09-08 09:56:30,833 INFO: Classification on Fake database for View2 with SVMLinear + +accuracy_score on train : 0.533333333333 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.514851485149 + - Score on test : 0.543209876543 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.514851485149 + - Score on test : 0.543209876543 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0654069940168 + - Score on test : 0.169618786115 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.52 + - Score on test : 0.55 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.509803921569 + - Score on test : 0.536585365854 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.532679738562 + - Score on test : 0.584619213539 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 +2016-09-08 09:56:30,833 INFO: Done: Result Analysis +2016-09-08 09:56:30,973 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:30,973 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:30,974 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 09:56:30,974 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 09:56:30,974 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:30,974 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:30,974 DEBUG: Info: Shape X_train:(210, 20), Length of y_train:210 +2016-09-08 09:56:30,974 DEBUG: Info: Shape X_train:(210, 20), Length of y_train:210 +2016-09-08 09:56:30,974 DEBUG: Info: Shape X_test:(90, 20), Length of y_test:90 +2016-09-08 09:56:30,974 DEBUG: Info: Shape X_test:(90, 20), Length of y_test:90 +2016-09-08 09:56:30,975 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:30,975 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:30,975 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:30,975 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:31,020 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:31,020 DEBUG: Start: Training +2016-09-08 09:56:31,024 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:31,024 DEBUG: Start: Training +2016-09-08 09:56:31,039 DEBUG: Info: Time for Training: 0.0657398700714[s] +2016-09-08 09:56:31,039 DEBUG: Done: Training +2016-09-08 09:56:31,039 DEBUG: Start: Predicting +2016-09-08 09:56:31,043 DEBUG: Info: Time for Training: 0.0697932243347[s] +2016-09-08 09:56:31,043 DEBUG: Done: Training +2016-09-08 09:56:31,043 DEBUG: Start: Predicting +2016-09-08 09:56:31,045 DEBUG: Done: Predicting +2016-09-08 09:56:31,045 DEBUG: Start: Getting Results +2016-09-08 09:56:31,047 DEBUG: Done: Predicting +2016-09-08 09:56:31,047 DEBUG: Start: Getting Results +2016-09-08 09:56:31,077 DEBUG: Done: Getting Results +2016-09-08 09:56:31,077 INFO: Classification on Fake database for View2 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588235294118 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588235294118 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0350036509618 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.448717948718 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.853658536585 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.488053758089 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:56:31,077 INFO: Done: Result Analysis +2016-09-08 09:56:31,077 DEBUG: Done: Getting Results +2016-09-08 09:56:31,077 INFO: Classification on Fake database for View2 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.453333333333 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.453333333333 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0695369227879 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.414634146341 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533847685416 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 +2016-09-08 09:56:31,078 INFO: Done: Result Analysis +2016-09-08 09:56:31,220 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:31,220 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:31,220 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:56:31,220 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:56:31,221 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:31,221 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:31,221 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:31,221 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:31,221 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:31,221 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:31,221 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:31,222 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:31,222 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:31,222 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:31,257 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:31,257 DEBUG: Start: Training +2016-09-08 09:56:31,259 DEBUG: Info: Time for Training: 0.0391139984131[s] +2016-09-08 09:56:31,259 DEBUG: Done: Training +2016-09-08 09:56:31,259 DEBUG: Start: Predicting +2016-09-08 09:56:31,262 DEBUG: Done: Predicting +2016-09-08 09:56:31,262 DEBUG: Start: Getting Results +2016-09-08 09:56:31,270 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:31,270 DEBUG: Start: Training +2016-09-08 09:56:31,274 DEBUG: Info: Time for Training: 0.0547120571136[s] +2016-09-08 09:56:31,275 DEBUG: Done: Training +2016-09-08 09:56:31,275 DEBUG: Start: Predicting +2016-09-08 09:56:31,277 DEBUG: Done: Predicting +2016-09-08 09:56:31,277 DEBUG: Start: Getting Results +2016-09-08 09:56:31,302 DEBUG: Done: Getting Results +2016-09-08 09:56:31,302 INFO: Classification on Fake database for View3 with DecisionTree + +accuracy_score on train : 0.985714285714 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.985714285714 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.985507246377 + - Score on test : 0.516853932584 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.985507246377 + - Score on test : 0.516853932584 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0142857142857 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.985714285714 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.971825315808 + - Score on test : 0.0506833064614 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.971428571429 + - Score on test : 0.479166666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.560975609756 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.986111111111 + - Score on test : 0.525385764062 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0142857142857 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 +2016-09-08 09:56:31,303 INFO: Done: Result Analysis +2016-09-08 09:56:31,313 DEBUG: Done: Getting Results +2016-09-08 09:56:31,314 INFO: Classification on Fake database for View3 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 10, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511627906977 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511627906977 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.066931612238 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.488888888889 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.536585365854 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533598805376 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 +2016-09-08 09:56:31,314 INFO: Done: Result Analysis +2016-09-08 09:56:31,468 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:31,468 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 09:56:31,468 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:31,468 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:31,469 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 09:56:31,469 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:31,469 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:31,469 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:31,469 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:31,469 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:31,469 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:31,469 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:31,470 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:31,470 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:31,503 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:31,503 DEBUG: Start: Training +2016-09-08 09:56:31,504 DEBUG: Info: Time for Training: 0.0364670753479[s] +2016-09-08 09:56:31,504 DEBUG: Done: Training +2016-09-08 09:56:31,504 DEBUG: Start: Predicting +2016-09-08 09:56:31,511 DEBUG: Done: Predicting +2016-09-08 09:56:31,511 DEBUG: Start: Getting Results +2016-09-08 09:56:31,565 DEBUG: Done: Getting Results +2016-09-08 09:56:31,565 INFO: Classification on Fake database for View3 with KNN + +accuracy_score on train : 0.533333333333 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 42 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.505050505051 + - Score on test : 0.543209876543 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.505050505051 + - Score on test : 0.543209876543 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0644812208979 + - Score on test : 0.169618786115 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.520833333333 + - Score on test : 0.55 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.490196078431 + - Score on test : 0.536585365854 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.532135076253 + - Score on test : 0.584619213539 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 +2016-09-08 09:56:31,565 INFO: Done: Result Analysis +2016-09-08 09:56:31,800 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:31,800 DEBUG: Start: Training +2016-09-08 09:56:31,848 DEBUG: Info: Time for Training: 0.380412101746[s] +2016-09-08 09:56:31,848 DEBUG: Done: Training +2016-09-08 09:56:31,848 DEBUG: Start: Predicting +2016-09-08 09:56:31,855 DEBUG: Done: Predicting +2016-09-08 09:56:31,855 DEBUG: Start: Getting Results +2016-09-08 09:56:31,882 DEBUG: Done: Getting Results +2016-09-08 09:56:31,882 INFO: Classification on Fake database for View3 with RandomForest + +accuracy_score on train : 0.995238095238 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 19, max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.99512195122 + - Score on test : 0.488372093023 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.99512195122 + - Score on test : 0.488372093023 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.990515975943 + - Score on test : 0.0223105374127 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.990291262136 + - Score on test : 0.466666666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.512195121951 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.99537037037 + - Score on test : 0.511199601792 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 09:56:31,882 INFO: Done: Result Analysis +2016-09-08 09:56:32,016 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:32,016 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:56:32,016 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 09:56:32,016 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 09:56:32,016 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:32,016 DEBUG: Start: Determine Train/Test split +2016-09-08 09:56:32,017 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:32,017 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:56:32,017 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:32,017 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:56:32,017 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:32,017 DEBUG: Done: Determine Train/Test split +2016-09-08 09:56:32,017 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:32,017 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:56:32,062 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:32,062 DEBUG: Start: Training +2016-09-08 09:56:32,062 DEBUG: Info: Time for Training: 0.04727602005[s] +2016-09-08 09:56:32,063 DEBUG: Done: Training +2016-09-08 09:56:32,063 DEBUG: Start: Predicting +2016-09-08 09:56:32,066 DEBUG: Done: RandomSearch best settings +2016-09-08 09:56:32,067 DEBUG: Start: Training +2016-09-08 09:56:32,075 DEBUG: Done: Predicting +2016-09-08 09:56:32,076 DEBUG: Start: Getting Results +2016-09-08 09:56:32,089 DEBUG: Info: Time for Training: 0.0742139816284[s] +2016-09-08 09:56:32,090 DEBUG: Done: Training +2016-09-08 09:56:32,090 DEBUG: Start: Predicting +2016-09-08 09:56:32,093 DEBUG: Done: Predicting +2016-09-08 09:56:32,093 DEBUG: Start: Getting Results +2016-09-08 09:56:32,100 DEBUG: Done: Getting Results +2016-09-08 09:56:32,100 INFO: Classification on Fake database for View3 with SGD + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : l1 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 09:56:32,100 INFO: Done: Result Analysis +2016-09-08 09:56:32,122 DEBUG: Done: Getting Results +2016-09-08 09:56:32,122 INFO: Classification on Fake database for View3 with SVMLinear + +accuracy_score on train : 0.552380952381 +accuracy_score on test : 0.566666666667 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.552380952381 + - Score on test : 0.566666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.520408163265 + - Score on test : 0.506329113924 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.520408163265 + - Score on test : 0.506329113924 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.447619047619 + - Score on test : 0.433333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.552380952381 + - Score on test : 0.566666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.10237359911 + - Score on test : 0.121459622637 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.542553191489 + - Score on test : 0.526315789474 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.487804878049 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.550925925926 + - Score on test : 0.560228969637 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.447619047619 + - Score on test : 0.433333333333 + + + Classification took 0:00:00 +2016-09-08 09:56:32,122 INFO: Done: Result Analysis +2016-09-08 09:56:32,424 INFO: ### Main Programm for Multiview Classification +2016-09-08 09:56:32,425 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 09:56:32,426 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 09:56:32,426 INFO: Info: Shape of View1 :(300, 19) +2016-09-08 09:56:32,427 INFO: Info: Shape of View2 :(300, 20) +2016-09-08 09:56:32,427 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 09:56:32,428 INFO: Done: Read Database Files +2016-09-08 09:56:32,428 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 09:56:32,431 INFO: ### Main Programm for Multiview Classification +2016-09-08 09:56:32,432 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 09:56:32,432 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 09:56:32,433 INFO: Done: Determine validation split +2016-09-08 09:56:32,433 INFO: Start: Determine 5 folds +2016-09-08 09:56:32,433 INFO: Info: Shape of View1 :(300, 19) +2016-09-08 09:56:32,434 INFO: Info: Shape of View2 :(300, 20) +2016-09-08 09:56:32,434 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 09:56:32,434 INFO: Done: Read Database Files +2016-09-08 09:56:32,434 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 09:56:32,439 INFO: Done: Determine validation split +2016-09-08 09:56:32,439 INFO: Start: Determine 5 folds +2016-09-08 09:56:32,445 INFO: Info: Length of Learning Sets: 169 +2016-09-08 09:56:32,445 INFO: Info: Length of Testing Sets: 42 +2016-09-08 09:56:32,445 INFO: Info: Length of Validation Set: 89 +2016-09-08 09:56:32,445 INFO: Done: Determine folds +2016-09-08 09:56:32,445 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 09:56:32,445 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 09:56:32,445 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:32,448 INFO: Info: Length of Learning Sets: 169 +2016-09-08 09:56:32,448 INFO: Info: Length of Testing Sets: 42 +2016-09-08 09:56:32,449 INFO: Info: Length of Validation Set: 89 +2016-09-08 09:56:32,449 INFO: Done: Determine folds +2016-09-08 09:56:32,449 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 09:56:32,449 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 09:56:32,449 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:32,500 DEBUG: Done: Random search for SGD +2016-09-08 09:56:32,501 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:32,504 DEBUG: Done: Random search for SGD +2016-09-08 09:56:32,504 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:32,551 DEBUG: Done: Random search for SGD +2016-09-08 09:56:32,551 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:32,556 DEBUG: Done: Random search for SGD +2016-09-08 09:56:32,556 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:32,600 DEBUG: Done: Random search for SGD +2016-09-08 09:56:32,600 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:32,607 DEBUG: Done: Random search for SGD +2016-09-08 09:56:32,607 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:32,651 DEBUG: Done: Random search for SGD +2016-09-08 09:56:32,660 DEBUG: Done: Random search for SGD +2016-09-08 09:56:32,717 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 09:56:32,717 INFO: Start: Classification +2016-09-08 09:56:32,717 INFO: Start: Fold number 1 +2016-09-08 09:56:32,745 INFO: Start: Classification +2016-09-08 09:56:32,751 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 09:56:32,751 INFO: Start: Classification +2016-09-08 09:56:32,751 INFO: Start: Fold number 1 +2016-09-08 09:56:32,771 INFO: Done: Fold number 1 +2016-09-08 09:56:32,772 INFO: Start: Fold number 2 +2016-09-08 09:56:32,778 INFO: Start: Classification +2016-09-08 09:56:32,798 INFO: Start: Classification +2016-09-08 09:56:32,824 INFO: Done: Fold number 2 +2016-09-08 09:56:32,824 INFO: Start: Fold number 3 +2016-09-08 09:56:32,847 INFO: Done: Fold number 1 +2016-09-08 09:56:32,848 INFO: Start: Fold number 2 +2016-09-08 09:56:32,850 INFO: Start: Classification +2016-09-08 09:56:32,874 INFO: Start: Classification +2016-09-08 09:56:32,876 INFO: Done: Fold number 3 +2016-09-08 09:56:32,877 INFO: Start: Fold number 4 +2016-09-08 09:56:32,903 INFO: Start: Classification +2016-09-08 09:56:32,928 INFO: Done: Fold number 4 +2016-09-08 09:56:32,928 INFO: Start: Fold number 5 +2016-09-08 09:56:32,944 INFO: Done: Fold number 2 +2016-09-08 09:56:32,944 INFO: Start: Fold number 3 +2016-09-08 09:56:32,955 INFO: Start: Classification +2016-09-08 09:56:32,971 INFO: Start: Classification +2016-09-08 09:56:32,981 INFO: Done: Fold number 5 +2016-09-08 09:56:32,981 INFO: Done: Classification +2016-09-08 09:56:32,981 INFO: Info: Time for Classification: 0[s] +2016-09-08 09:56:32,981 INFO: Start: Result Analysis for Fusion +2016-09-08 09:56:33,041 INFO: Done: Fold number 3 +2016-09-08 09:56:33,041 INFO: Start: Fold number 4 +2016-09-08 09:56:33,068 INFO: Start: Classification +2016-09-08 09:56:33,109 INFO: Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 64.9704142012 + -On Test : 42.380952381 + -On Validation : 48.7640449438 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Bayesian Inference using a weight for each view : 0.241839908393, 0.362121620258, 0.0533308084229, 0.342707662926 + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : elasticnet + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : modified_huber, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 09:56:33,109 INFO: Done: Result Analysis +2016-09-08 09:56:33,137 INFO: Done: Fold number 4 +2016-09-08 09:56:33,137 INFO: Start: Fold number 5 +2016-09-08 09:56:33,163 INFO: Start: Classification +2016-09-08 09:56:33,229 INFO: Done: Fold number 5 +2016-09-08 09:56:33,229 INFO: Done: Classification +2016-09-08 09:56:33,229 INFO: Info: Time for Classification: 0[s] +2016-09-08 09:56:33,229 INFO: Start: Result Analysis for Fusion +2016-09-08 09:56:33,351 INFO: Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 62.2485207101 + -On Test : 47.619047619 + -On Validation : 52.1348314607 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Majority Voting + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : elasticnet + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : modified_huber, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 09:56:33,351 INFO: Done: Result Analysis +2016-09-08 09:56:33,471 INFO: ### Main Programm for Multiview Classification +2016-09-08 09:56:33,471 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 09:56:33,472 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 09:56:33,472 INFO: Info: Shape of View1 :(300, 19) +2016-09-08 09:56:33,473 INFO: Info: Shape of View2 :(300, 20) +2016-09-08 09:56:33,473 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 09:56:33,473 INFO: Done: Read Database Files +2016-09-08 09:56:33,474 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 09:56:33,478 INFO: ### Main Programm for Multiview Classification +2016-09-08 09:56:33,478 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 09:56:33,479 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 09:56:33,480 INFO: Info: Shape of View1 :(300, 19) +2016-09-08 09:56:33,481 INFO: Done: Determine validation split +2016-09-08 09:56:33,481 INFO: Start: Determine 5 folds +2016-09-08 09:56:33,481 INFO: Info: Shape of View2 :(300, 20) +2016-09-08 09:56:33,482 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 09:56:33,482 INFO: Done: Read Database Files +2016-09-08 09:56:33,482 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 09:56:33,485 INFO: Done: Determine validation split +2016-09-08 09:56:33,485 INFO: Start: Determine 5 folds +2016-09-08 09:56:33,488 INFO: Info: Length of Learning Sets: 169 +2016-09-08 09:56:33,489 INFO: Info: Length of Testing Sets: 42 +2016-09-08 09:56:33,489 INFO: Info: Length of Validation Set: 89 +2016-09-08 09:56:33,489 INFO: Done: Determine folds +2016-09-08 09:56:33,489 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 09:56:33,489 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 09:56:33,489 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:33,495 INFO: Info: Length of Learning Sets: 169 +2016-09-08 09:56:33,495 INFO: Info: Length of Testing Sets: 42 +2016-09-08 09:56:33,495 INFO: Info: Length of Validation Set: 89 +2016-09-08 09:56:33,495 INFO: Done: Determine folds +2016-09-08 09:56:33,496 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 09:56:33,496 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 09:56:33,496 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:33,544 DEBUG: Done: Random search for SGD +2016-09-08 09:56:33,545 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:33,550 DEBUG: Done: Random search for SGD +2016-09-08 09:56:33,550 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:33,596 DEBUG: Done: Random search for SGD +2016-09-08 09:56:33,596 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:33,600 DEBUG: Done: Random search for SGD +2016-09-08 09:56:33,601 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:33,646 DEBUG: Done: Random search for SGD +2016-09-08 09:56:33,646 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:33,650 DEBUG: Done: Random search for SGD +2016-09-08 09:56:33,650 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 09:56:33,698 DEBUG: Done: Random search for SGD +2016-09-08 09:56:33,698 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 09:56:33,698 INFO: Start: Classification +2016-09-08 09:56:33,699 INFO: Start: Fold number 1 +2016-09-08 09:56:33,702 DEBUG: Done: Random search for SGD +2016-09-08 09:56:33,760 INFO: Start: Classification +2016-09-08 09:56:33,767 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 09:56:33,767 INFO: Start: Classification +2016-09-08 09:56:33,767 INFO: Start: Fold number 1 +2016-09-08 09:56:33,790 INFO: Done: Fold number 1 +2016-09-08 09:56:33,790 INFO: Start: Fold number 2 +2016-09-08 09:56:33,794 INFO: Start: Classification +2016-09-08 09:56:33,825 INFO: Done: Fold number 1 +2016-09-08 09:56:33,825 INFO: Start: Fold number 2 +2016-09-08 09:56:33,835 INFO: Start: Classification +2016-09-08 09:56:33,851 INFO: Start: Classification +2016-09-08 09:56:33,864 INFO: Done: Fold number 2 +2016-09-08 09:56:33,865 INFO: Start: Fold number 3 +2016-09-08 09:56:33,882 INFO: Done: Fold number 2 +2016-09-08 09:56:33,882 INFO: Start: Fold number 3 +2016-09-08 09:56:33,908 INFO: Start: Classification +2016-09-08 09:56:33,909 INFO: Start: Classification +2016-09-08 09:56:33,938 INFO: Done: Fold number 3 +2016-09-08 09:56:33,938 INFO: Start: Fold number 4 +2016-09-08 09:56:33,938 INFO: Done: Fold number 3 +2016-09-08 09:56:33,939 INFO: Start: Fold number 4 +2016-09-08 09:56:33,965 INFO: Start: Classification +2016-09-08 09:56:33,984 INFO: Start: Classification +2016-09-08 09:56:33,997 INFO: Done: Fold number 4 +2016-09-08 09:56:33,997 INFO: Start: Fold number 5 +2016-09-08 09:56:34,014 INFO: Done: Fold number 4 +2016-09-08 09:56:34,014 INFO: Start: Fold number 5 +2016-09-08 09:56:34,024 INFO: Start: Classification +2016-09-08 09:56:34,054 INFO: Done: Fold number 5 +2016-09-08 09:56:34,054 INFO: Done: Classification +2016-09-08 09:56:34,054 INFO: Info: Time for Classification: 0[s] +2016-09-08 09:56:34,055 INFO: Start: Result Analysis for Fusion +2016-09-08 09:56:34,059 INFO: Start: Classification +2016-09-08 09:56:34,089 INFO: Done: Fold number 5 +2016-09-08 09:56:34,089 INFO: Done: Classification +2016-09-08 09:56:34,089 INFO: Info: Time for Classification: 0[s] +2016-09-08 09:56:34,089 INFO: Start: Result Analysis for Fusion +2016-09-08 09:56:34,221 INFO: Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 63.4319526627 + -On Test : 47.619047619 + -On Validation : 44.2696629213 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with SVM for linear + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : elasticnet + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : modified_huber, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 09:56:34,222 INFO: Done: Result Analysis diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..03f454c1907151b2cd4700f0b9fa21f634a895a3 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View0 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.377777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 10, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.377777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.622222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.377777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.249628898234 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.325581395349 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.341463414634 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.37481333997 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.622222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..1121217156862cbd341038acd8d0bb7420aea4f5 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.377777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.377777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.348837209302 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.348837209302 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.622222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.377777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.245415911539 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.365853658537 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.376804380289 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.622222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..717561a91111c362546c229613b377405c534f58 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095627Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with KNN + +accuracy_score on train : 0.609523809524 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 42 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.609523809524 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.467532467532 + - Score on test : 0.212121212121 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.467532467532 + - Score on test : 0.212121212121 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.390476190476 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.609523809524 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.237135686833 + - Score on test : -0.218615245335 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.692307692308 + - Score on test : 0.28 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.352941176471 + - Score on test : 0.170731707317 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.602396514161 + - Score on test : 0.401692384271 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.390476190476 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..db639d874780820a793d5b425131f599b619d1a1 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View1 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 10, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.119960179194 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.440019910403 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..63f60667fbbb850f1c17b27d456a7ad94bf2a1cd --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with DecisionTree + +accuracy_score on train : 0.957142857143 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.957142857143 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.957746478873 + - Score on test : 0.444444444444 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.957746478873 + - Score on test : 0.444444444444 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0428571428571 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.957142857143 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.917792101918 + - Score on test : -0.104031856645 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.918918918919 + - Score on test : 0.408163265306 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.487804878049 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.958333333333 + - Score on test : 0.447984071677 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0428571428571 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..450f0ea162ee5681ba44e7e36c2da99ff7c208db --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with KNN + +accuracy_score on train : 0.561904761905 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 42 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.561904761905 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.6 + - Score on test : 0.559139784946 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.6 + - Score on test : 0.559139784946 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.438095238095 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.561904761905 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.133359904768 + - Score on test : 0.104395047556 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.5390625 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.676470588235 + - Score on test : 0.634146341463 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.565087145969 + - Score on test : 0.551767048283 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.438095238095 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..3d2d8158208053f14d8d289f1b53da5a0ff267a7 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 19, max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.423529411765 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.423529411765 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0912478416452 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.409090909091 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.439024390244 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.454206072673 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..511f2feec38b0ab80eca16401cbe7d50a5a52472 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SGD + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : l1 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..2bda3cf3e9865595e5900f4a617e2f7210d1b111 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMLinear + +accuracy_score on train : 0.52380952381 +accuracy_score on test : 0.477777777778 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.52380952381 + - Score on test : 0.477777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.532710280374 + - Score on test : 0.459770114943 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.532710280374 + - Score on test : 0.459770114943 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.47619047619 + - Score on test : 0.522222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.52380952381 + - Score on test : 0.477777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0496545019224 + - Score on test : -0.0426484477255 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.508928571429 + - Score on test : 0.434782608696 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.558823529412 + - Score on test : 0.487804878049 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.524782135076 + - Score on test : 0.478596316575 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.47619047619 + - Score on test : 0.522222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..59cdb6c697d146d7d3ed98f76ae3654d60df8f00 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.475 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.475 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0555284586866 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.487179487179 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.463414634146 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.52762568442 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..a96b05b6ca83ab7a200d3d56ec086a92a28bbee4 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095628Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.409090909091 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.409090909091 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.152357995542 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.382978723404 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.439024390244 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.423593827775 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..46f9acd1a6f415e8ac5c14de01639553c10c8914 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with RandomForest + +accuracy_score on train : 0.995238095238 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 19, max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.995073891626 + - Score on test : 0.516853932584 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.995073891626 + - Score on test : 0.516853932584 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.990510833227 + - Score on test : 0.0506833064614 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.479166666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.990196078431 + - Score on test : 0.560975609756 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.995098039216 + - Score on test : 0.525385764062 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..c2daea4e1eb4967649e0a7bd2c099f23f75f51e8 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SGD + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : l1 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..65a1d806d6c532ff7273d35788ad31ff125bf076 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMLinear + +accuracy_score on train : 0.514285714286 +accuracy_score on test : 0.577777777778 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.514285714286 + - Score on test : 0.577777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.514285714286 + - Score on test : 0.547619047619 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.514285714286 + - Score on test : 0.547619047619 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.485714285714 + - Score on test : 0.422222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.514285714286 + - Score on test : 0.577777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0294117647059 + - Score on test : 0.152357995542 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.53488372093 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.529411764706 + - Score on test : 0.560975609756 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.514705882353 + - Score on test : 0.576406172225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.485714285714 + - Score on test : 0.422222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..ffa02970261e5fd6c6ee7c7c89f28a71c8c1cb80 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.394366197183 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.394366197183 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0157759322964 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.466666666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.341463414634 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.507466401195 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..664dd67dd1b310bf15826196f3f32bed9c3db30e --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095629Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 19) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0193709711057 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.44 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.268292682927 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.491289198606 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..1026306a026375c58d260dba33e7e4dd921b4d8c --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View2 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 10, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.375 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.375 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.124563839757 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.384615384615 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.365853658537 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.438028870085 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..e47d02bff46273cea19d86bfd2d436741ac501b5 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.433333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.385542168675 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.385542168675 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.140124435511 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.380952380952 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.390243902439 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.429815828771 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..15b07a9ccd577250db3514055248e247ee8ca5ac --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with KNN + +accuracy_score on train : 0.542857142857 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 42 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.542857142857 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.323943661972 + - Score on test : 0.327272727273 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.323943661972 + - Score on test : 0.327272727273 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.457142857143 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.542857142857 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0866552427925 + - Score on test : 0.161417724438 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.575 + - Score on test : 0.642857142857 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.225490196078 + - Score on test : 0.219512195122 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.534041394336 + - Score on test : 0.558735689398 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.457142857143 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..70517c8c47fe2344c26b87465cad2ea40f5e082b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 19, max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.55421686747 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.55421686747 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.172919516163 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.547619047619 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.560975609756 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.586610253858 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..fd86ceca50e091c5d36f8d06c22854be9865701e --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SGD + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : l1 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..ea0b1413c2f836e668154e840087b108a965c0cf --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095630Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMLinear + +accuracy_score on train : 0.533333333333 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.514851485149 + - Score on test : 0.543209876543 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.514851485149 + - Score on test : 0.543209876543 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0654069940168 + - Score on test : 0.169618786115 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.52 + - Score on test : 0.55 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.509803921569 + - Score on test : 0.536585365854 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.532679738562 + - Score on test : 0.584619213539 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..7ea3a02f3a5aa6f8340cf74af6bcb60cdc9941ac --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View3 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 10, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511627906977 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511627906977 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.066931612238 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.488888888889 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.536585365854 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533598805376 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..a3cc29473420a6a6e84a2cba2dca5272ed452636 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with DecisionTree + +accuracy_score on train : 0.985714285714 +accuracy_score on test : 0.522222222222 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.985714285714 + - Score on test : 0.522222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.985507246377 + - Score on test : 0.516853932584 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.985507246377 + - Score on test : 0.516853932584 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0142857142857 + - Score on test : 0.477777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.985714285714 + - Score on test : 0.522222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.971825315808 + - Score on test : 0.0506833064614 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.971428571429 + - Score on test : 0.479166666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.560975609756 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.986111111111 + - Score on test : 0.525385764062 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0142857142857 + - Score on test : 0.477777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..889e2b6cb95717c6978b1c1df26d27a2cf59c58b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with KNN + +accuracy_score on train : 0.533333333333 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 42 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.505050505051 + - Score on test : 0.543209876543 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.505050505051 + - Score on test : 0.543209876543 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0644812208979 + - Score on test : 0.169618786115 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.520833333333 + - Score on test : 0.55 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.490196078431 + - Score on test : 0.536585365854 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.532135076253 + - Score on test : 0.584619213539 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.466666666667 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..df0c84da34a8bbfff60cea62bfe7affe7827d513 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with RandomForest + +accuracy_score on train : 0.995238095238 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 19, max_depth : 10 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.99512195122 + - Score on test : 0.488372093023 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.99512195122 + - Score on test : 0.488372093023 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.990515975943 + - Score on test : 0.0223105374127 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.990291262136 + - Score on test : 0.466666666667 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.512195121951 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.99537037037 + - Score on test : 0.511199601792 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..58bf957cbce20026512b4f37e8e68de25eb1edc6 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588235294118 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588235294118 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0350036509618 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.448717948718 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.853658536585 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.488053758089 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..b171204485ec1e649cf7ce79673da7cf18354e76 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095631Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 20) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.453333333333 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.453333333333 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0695369227879 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.414634146341 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533847685416 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095632Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095632Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..ac66b07e027445e598fa09e5c12feee404ee19f5 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095632Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with SGD + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : log, penalty : l1 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.653846153846 + - Score on test : 0.625954198473 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.455555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 1.0 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095632Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095632Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..c4a246c96517e4894fffb96593914dc482cc196e --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095632Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with SVMLinear + +accuracy_score on train : 0.552380952381 +accuracy_score on test : 0.566666666667 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 8627 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.552380952381 + - Score on test : 0.566666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.520408163265 + - Score on test : 0.506329113924 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.520408163265 + - Score on test : 0.506329113924 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.447619047619 + - Score on test : 0.433333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.552380952381 + - Score on test : 0.566666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.10237359911 + - Score on test : 0.121459622637 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.542553191489 + - Score on test : 0.526315789474 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.5 + - Score on test : 0.487804878049 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.550925925926 + - Score on test : 0.560228969637 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.447619047619 + - Score on test : 0.433333333333 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095633Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095633Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..5ec167846b7f8da92aa16cc60fec7b475c81e8a8 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095633Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,32 @@ + Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 64.9704142012 + -On Test : 42.380952381 + -On Validation : 48.7640449438 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Bayesian Inference using a weight for each view : 0.241839908393, 0.362121620258, 0.0533308084229, 0.342707662926 + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : elasticnet + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : modified_huber, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095633Results-Fusion-LateFusion-MajorityVoting-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095633Results-Fusion-LateFusion-MajorityVoting-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..f9c3ded12ed06616f706139519336a54251ed8d1 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095633Results-Fusion-LateFusion-MajorityVoting-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,32 @@ + Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 62.2485207101 + -On Test : 47.619047619 + -On Validation : 52.1348314607 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Majority Voting + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : elasticnet + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : modified_huber, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095634Results-Fusion-LateFusion-SVMForLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-095634Results-Fusion-LateFusion-SVMForLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..457213b07a7b6ea964f2fc98f31d8784d8005e60 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095634Results-Fusion-LateFusion-SVMForLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,32 @@ + Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 63.4319526627 + -On Test : 47.619047619 + -On Validation : 44.2696629213 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with SVM for linear + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : elasticnet + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : modified_huber, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095845-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log b/Code/MonoMutliViewClassifiers/Results/20160908-095845-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log new file mode 100644 index 0000000000000000000000000000000000000000..d434781c220fac77883c4cff413c638b54560262 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095845-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log @@ -0,0 +1,32 @@ +2016-09-08 09:58:45,489 DEBUG: Start: Creating 2 temporary datasets for multiprocessing +2016-09-08 09:58:45,489 WARNING: WARNING : /!\ This may use a lot of HDD storage space : 0.00010759375 Gbytes /!\ +2016-09-08 09:58:50,503 DEBUG: Start: Creating datasets for multiprocessing +2016-09-08 09:58:50,507 INFO: Start: Finding all available mono- & multiview algorithms +2016-09-08 09:58:50,558 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:58:50,558 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 09:58:50,559 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 09:58:50,559 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 09:58:50,559 DEBUG: Start: Determine Train/Test split +2016-09-08 09:58:50,559 DEBUG: Start: Determine Train/Test split +2016-09-08 09:58:50,559 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:58:50,559 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 09:58:50,559 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:58:50,559 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 09:58:50,560 DEBUG: Done: Determine Train/Test split +2016-09-08 09:58:50,560 DEBUG: Done: Determine Train/Test split +2016-09-08 09:58:50,560 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:58:50,560 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 09:58:50,594 DEBUG: Done: RandomSearch best settings +2016-09-08 09:58:50,595 DEBUG: Start: Training +2016-09-08 09:58:50,596 DEBUG: Info: Time for Training: 0.0384030342102[s] +2016-09-08 09:58:50,596 DEBUG: Done: Training +2016-09-08 09:58:50,596 DEBUG: Start: Predicting +2016-09-08 09:58:50,599 DEBUG: Done: Predicting +2016-09-08 09:58:50,599 DEBUG: Start: Getting Results +2016-09-08 09:58:50,609 DEBUG: Done: RandomSearch best settings +2016-09-08 09:58:50,609 DEBUG: Start: Training +2016-09-08 09:58:50,613 DEBUG: Info: Time for Training: 0.0557579994202[s] +2016-09-08 09:58:50,614 DEBUG: Done: Training +2016-09-08 09:58:50,614 DEBUG: Start: Predicting +2016-09-08 09:58:50,616 DEBUG: Done: Predicting +2016-09-08 09:58:50,617 DEBUG: Start: Getting Results diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-095958-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log b/Code/MonoMutliViewClassifiers/Results/20160908-095958-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log new file mode 100644 index 0000000000000000000000000000000000000000..9a77052c7e7b44b976d23973f6729d2cf0291fab --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-095958-CMultiV-Benchmark-Methyl_MiRNA__RNASeq_Clinic-Fake-LOG.log @@ -0,0 +1,2842 @@ +2016-09-08 09:59:58,949 DEBUG: Start: Creating 2 temporary datasets for multiprocessing +2016-09-08 09:59:58,950 WARNING: WARNING : /!\ This may use a lot of HDD storage space : 0.000152125 Gbytes /!\ +2016-09-08 10:00:03,964 DEBUG: Start: Creating datasets for multiprocessing +2016-09-08 10:00:03,968 INFO: Start: Finding all available mono- & multiview algorithms +2016-09-08 10:00:04,015 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:04,015 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:04,015 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 10:00:04,015 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 10:00:04,016 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:04,016 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:04,016 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:04,016 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:04,017 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:04,017 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:04,017 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:04,017 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:04,017 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:04,017 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:04,051 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:04,052 DEBUG: Start: Training +2016-09-08 10:00:04,053 DEBUG: Info: Time for Training: 0.0387499332428[s] +2016-09-08 10:00:04,054 DEBUG: Done: Training +2016-09-08 10:00:04,054 DEBUG: Start: Predicting +2016-09-08 10:00:04,056 DEBUG: Done: Predicting +2016-09-08 10:00:04,056 DEBUG: Start: Getting Results +2016-09-08 10:00:04,067 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:04,067 DEBUG: Start: Training +2016-09-08 10:00:04,071 DEBUG: Info: Time for Training: 0.0559167861938[s] +2016-09-08 10:00:04,071 DEBUG: Done: Training +2016-09-08 10:00:04,071 DEBUG: Start: Predicting +2016-09-08 10:00:04,073 DEBUG: Done: Predicting +2016-09-08 10:00:04,074 DEBUG: Start: Getting Results +2016-09-08 10:00:04,106 DEBUG: Done: Getting Results +2016-09-08 10:00:04,106 INFO: Classification on Fake database for View0 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.56862745098 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.56862745098 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.00503027272866 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.557692307692 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.58 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5025 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 10:00:04,106 INFO: Done: Result Analysis +2016-09-08 10:00:04,108 DEBUG: Done: Getting Results +2016-09-08 10:00:04,108 INFO: Classification on Fake database for View0 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 9, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.576923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.576923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -2.5337258102e-17 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.555555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.6 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 10:00:04,109 INFO: Done: Result Analysis +2016-09-08 10:00:04,267 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:04,268 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 10:00:04,268 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:04,268 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:04,268 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 10:00:04,268 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:04,269 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:04,269 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:04,269 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:04,269 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:04,269 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:04,269 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:04,269 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:04,270 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:04,300 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:04,301 DEBUG: Start: Training +2016-09-08 10:00:04,301 DEBUG: Info: Time for Training: 0.0347349643707[s] +2016-09-08 10:00:04,301 DEBUG: Done: Training +2016-09-08 10:00:04,301 DEBUG: Start: Predicting +2016-09-08 10:00:04,307 DEBUG: Done: Predicting +2016-09-08 10:00:04,307 DEBUG: Start: Getting Results +2016-09-08 10:00:04,347 DEBUG: Done: Getting Results +2016-09-08 10:00:04,347 INFO: Classification on Fake database for View0 with KNN + +accuracy_score on train : 0.580952380952 +accuracy_score on test : 0.6 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 24 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.580952380952 + - Score on test : 0.6 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.661538461538 + - Score on test : 0.678571428571 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.661538461538 + - Score on test : 0.678571428571 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.419047619048 + - Score on test : 0.4 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.580952380952 + - Score on test : 0.6 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.130162282504 + - Score on test : 0.171735516296 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.601398601399 + - Score on test : 0.612903225806 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.735042735043 + - Score on test : 0.76 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.561069754618 + - Score on test : 0.58 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.419047619048 + - Score on test : 0.4 + + + Classification took 0:00:00 +2016-09-08 10:00:04,347 INFO: Done: Result Analysis +2016-09-08 10:00:04,669 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:04,669 DEBUG: Start: Training +2016-09-08 10:00:04,730 DEBUG: Info: Time for Training: 0.462595939636[s] +2016-09-08 10:00:04,730 DEBUG: Done: Training +2016-09-08 10:00:04,730 DEBUG: Start: Predicting +2016-09-08 10:00:04,737 DEBUG: Done: Predicting +2016-09-08 10:00:04,737 DEBUG: Start: Getting Results +2016-09-08 10:00:04,771 DEBUG: Done: Getting Results +2016-09-08 10:00:04,771 INFO: Classification on Fake database for View0 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 24, max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.52427184466 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.52427184466 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.111088444626 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.509433962264 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.54 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.445 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 10:00:04,771 INFO: Done: Result Analysis +2016-09-08 10:00:04,916 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:04,916 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 10:00:04,917 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:04,917 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:04,917 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 10:00:04,917 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:04,918 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:04,918 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:04,918 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:04,918 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:04,919 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:04,919 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:04,919 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:04,919 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:04,964 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:04,964 DEBUG: Start: Training +2016-09-08 10:00:04,965 DEBUG: Info: Time for Training: 0.049379825592[s] +2016-09-08 10:00:04,965 DEBUG: Done: Training +2016-09-08 10:00:04,965 DEBUG: Start: Predicting +2016-09-08 10:00:04,971 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:04,971 DEBUG: Start: Training +2016-09-08 10:00:04,979 DEBUG: Done: Predicting +2016-09-08 10:00:04,980 DEBUG: Start: Getting Results +2016-09-08 10:00:04,997 DEBUG: Info: Time for Training: 0.0806729793549[s] +2016-09-08 10:00:04,997 DEBUG: Done: Training +2016-09-08 10:00:04,997 DEBUG: Start: Predicting +2016-09-08 10:00:05,000 DEBUG: Done: Predicting +2016-09-08 10:00:05,000 DEBUG: Start: Getting Results +2016-09-08 10:00:05,005 DEBUG: Done: Getting Results +2016-09-08 10:00:05,005 INFO: Classification on Fake database for View0 with SGD + +accuracy_score on train : 0.57619047619 +accuracy_score on test : 0.644444444444 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : modified_huber, penalty : l2 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.57619047619 + - Score on test : 0.644444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.661596958175 + - Score on test : 0.724137931034 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.661596958175 + - Score on test : 0.724137931034 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.42380952381 + - Score on test : 0.355555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.57619047619 + - Score on test : 0.644444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.117819078215 + - Score on test : 0.269679944985 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.595890410959 + - Score on test : 0.636363636364 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.74358974359 + - Score on test : 0.84 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.55459057072 + - Score on test : 0.62 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.42380952381 + - Score on test : 0.355555555556 + + + Classification took 0:00:00 +2016-09-08 10:00:05,005 INFO: Done: Result Analysis +2016-09-08 10:00:05,029 DEBUG: Done: Getting Results +2016-09-08 10:00:05,029 INFO: Classification on Fake database for View0 with SVMLinear + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.609523809524 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.609523809524 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0643090141189 + - Score on test : 0.0662541348869 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.581818181818 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.615384615385 + - Score on test : 0.64 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.468982630273 + - Score on test : 0.5325 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 +2016-09-08 10:00:05,030 INFO: Done: Result Analysis +2016-09-08 10:00:05,163 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:05,164 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 10:00:05,164 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:05,164 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:05,164 DEBUG: ### Classification - Database:Fake Feature:View0 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 10:00:05,164 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:05,164 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:05,165 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:05,165 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:05,165 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:05,165 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:05,165 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:05,165 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:05,165 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:05,212 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:05,213 DEBUG: Start: Training +2016-09-08 10:00:05,218 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:05,219 DEBUG: Start: Training +2016-09-08 10:00:05,230 DEBUG: Info: Time for Training: 0.0661790370941[s] +2016-09-08 10:00:05,230 DEBUG: Done: Training +2016-09-08 10:00:05,230 DEBUG: Start: Predicting +2016-09-08 10:00:05,235 DEBUG: Done: Predicting +2016-09-08 10:00:05,236 DEBUG: Start: Getting Results +2016-09-08 10:00:05,239 DEBUG: Info: Time for Training: 0.0759570598602[s] +2016-09-08 10:00:05,239 DEBUG: Done: Training +2016-09-08 10:00:05,239 DEBUG: Start: Predicting +2016-09-08 10:00:05,243 DEBUG: Done: Predicting +2016-09-08 10:00:05,243 DEBUG: Start: Getting Results +2016-09-08 10:00:05,275 DEBUG: Done: Getting Results +2016-09-08 10:00:05,275 INFO: Classification on Fake database for View0 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.6 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.6 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.714285714286 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.714285714286 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.4 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.6 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.171377655346 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.592105263158 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.9 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5625 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.4 + + + Classification took 0:00:00 +2016-09-08 10:00:05,275 INFO: Done: Result Analysis +2016-09-08 10:00:05,283 DEBUG: Done: Getting Results +2016-09-08 10:00:05,283 INFO: Classification on Fake database for View0 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0973655073258 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.617647058824 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.42 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5475 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 +2016-09-08 10:00:05,283 INFO: Done: Result Analysis +2016-09-08 10:00:05,410 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:05,410 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:05,411 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 10:00:05,411 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 10:00:05,411 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:05,411 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:05,411 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 10:00:05,411 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 10:00:05,411 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 10:00:05,411 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 10:00:05,412 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:05,412 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:05,412 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:05,412 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:05,447 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:05,447 DEBUG: Start: Training +2016-09-08 10:00:05,450 DEBUG: Info: Time for Training: 0.0400369167328[s] +2016-09-08 10:00:05,450 DEBUG: Done: Training +2016-09-08 10:00:05,450 DEBUG: Start: Predicting +2016-09-08 10:00:05,452 DEBUG: Done: Predicting +2016-09-08 10:00:05,453 DEBUG: Start: Getting Results +2016-09-08 10:00:05,462 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:05,462 DEBUG: Start: Training +2016-09-08 10:00:05,467 DEBUG: Info: Time for Training: 0.0573270320892[s] +2016-09-08 10:00:05,467 DEBUG: Done: Training +2016-09-08 10:00:05,467 DEBUG: Start: Predicting +2016-09-08 10:00:05,471 DEBUG: Done: Predicting +2016-09-08 10:00:05,471 DEBUG: Start: Getting Results +2016-09-08 10:00:05,499 DEBUG: Done: Getting Results +2016-09-08 10:00:05,499 INFO: Classification on Fake database for View1 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.637168141593 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.637168141593 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0487950036474 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.571428571429 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.72 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 +2016-09-08 10:00:05,499 INFO: Done: Result Analysis +2016-09-08 10:00:05,516 DEBUG: Done: Getting Results +2016-09-08 10:00:05,516 INFO: Classification on Fake database for View1 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.555555555556 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 9, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.555555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.655172413793 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.655172413793 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.444444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.555555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0674199862463 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.575757575758 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.76 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.53 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.444444444444 + + + Classification took 0:00:00 +2016-09-08 10:00:05,517 INFO: Done: Result Analysis +2016-09-08 10:00:05,656 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:05,656 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 10:00:05,656 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:05,656 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:05,656 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 10:00:05,657 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:05,657 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 10:00:05,657 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 10:00:05,657 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 10:00:05,657 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:05,657 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 10:00:05,657 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:05,657 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:05,657 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:05,688 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:05,688 DEBUG: Start: Training +2016-09-08 10:00:05,689 DEBUG: Info: Time for Training: 0.0336298942566[s] +2016-09-08 10:00:05,689 DEBUG: Done: Training +2016-09-08 10:00:05,689 DEBUG: Start: Predicting +2016-09-08 10:00:05,695 DEBUG: Done: Predicting +2016-09-08 10:00:05,695 DEBUG: Start: Getting Results +2016-09-08 10:00:05,735 DEBUG: Done: Getting Results +2016-09-08 10:00:05,735 INFO: Classification on Fake database for View1 with KNN + +accuracy_score on train : 0.566666666667 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 24 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.566666666667 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.659176029963 + - Score on test : 0.542056074766 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.659176029963 + - Score on test : 0.542056074766 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.433333333333 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.566666666667 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0939782359481 + - Score on test : -0.123737644978 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.586666666667 + - Score on test : 0.508771929825 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.752136752137 + - Score on test : 0.58 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.542735042735 + - Score on test : 0.44 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.433333333333 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 10:00:05,735 INFO: Done: Result Analysis +2016-09-08 10:00:06,049 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:06,050 DEBUG: Start: Training +2016-09-08 10:00:06,110 DEBUG: Info: Time for Training: 0.454261064529[s] +2016-09-08 10:00:06,110 DEBUG: Done: Training +2016-09-08 10:00:06,110 DEBUG: Start: Predicting +2016-09-08 10:00:06,117 DEBUG: Done: Predicting +2016-09-08 10:00:06,117 DEBUG: Start: Getting Results +2016-09-08 10:00:06,151 DEBUG: Done: Getting Results +2016-09-08 10:00:06,151 INFO: Classification on Fake database for View1 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.566666666667 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 24, max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.566666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.621359223301 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.621359223301 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.433333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.566666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.116137919381 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.603773584906 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.64 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5575 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.433333333333 + + + Classification took 0:00:00 +2016-09-08 10:00:06,151 INFO: Done: Result Analysis +2016-09-08 10:00:06,304 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:06,304 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 10:00:06,304 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:06,304 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:06,305 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 10:00:06,305 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:06,305 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 10:00:06,305 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 10:00:06,305 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:06,305 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:06,305 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 10:00:06,306 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 10:00:06,306 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:06,306 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:06,348 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:06,349 DEBUG: Start: Training +2016-09-08 10:00:06,349 DEBUG: Info: Time for Training: 0.0459690093994[s] +2016-09-08 10:00:06,349 DEBUG: Done: Training +2016-09-08 10:00:06,350 DEBUG: Start: Predicting +2016-09-08 10:00:06,356 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:06,356 DEBUG: Start: Training +2016-09-08 10:00:06,379 DEBUG: Done: Predicting +2016-09-08 10:00:06,379 DEBUG: Start: Getting Results +2016-09-08 10:00:06,381 DEBUG: Info: Time for Training: 0.0768840312958[s] +2016-09-08 10:00:06,381 DEBUG: Done: Training +2016-09-08 10:00:06,381 DEBUG: Start: Predicting +2016-09-08 10:00:06,387 DEBUG: Done: Predicting +2016-09-08 10:00:06,387 DEBUG: Start: Getting Results +2016-09-08 10:00:06,406 DEBUG: Done: Getting Results +2016-09-08 10:00:06,406 INFO: Classification on Fake database for View1 with SGD + +accuracy_score on train : 0.566666666667 +accuracy_score on test : 0.555555555556 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : modified_huber, penalty : l2 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.566666666667 + - Score on test : 0.555555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.640316205534 + - Score on test : 0.666666666667 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.640316205534 + - Score on test : 0.666666666667 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.433333333333 + - Score on test : 0.444444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.566666666667 + - Score on test : 0.555555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.104925880155 + - Score on test : 0.0597614304667 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.595588235294 + - Score on test : 0.571428571429 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.692307692308 + - Score on test : 0.8 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.550454921423 + - Score on test : 0.525 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.433333333333 + - Score on test : 0.444444444444 + + + Classification took 0:00:00 +2016-09-08 10:00:06,407 INFO: Done: Result Analysis +2016-09-08 10:00:06,420 DEBUG: Done: Getting Results +2016-09-08 10:00:06,420 INFO: Classification on Fake database for View1 with SVMLinear + +accuracy_score on train : 0.490476190476 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.490476190476 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.566801619433 + - Score on test : 0.576923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.566801619433 + - Score on test : 0.576923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.509523809524 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.490476190476 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0479423260647 + - Score on test : -3.54721613428e-17 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.538461538462 + - Score on test : 0.555555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.598290598291 + - Score on test : 0.6 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.476564653984 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.509523809524 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 10:00:06,420 INFO: Done: Result Analysis +2016-09-08 10:00:06,547 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:06,547 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:06,547 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 10:00:06,548 DEBUG: ### Classification - Database:Fake Feature:View1 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 10:00:06,548 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:06,548 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:06,548 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 10:00:06,548 DEBUG: Info: Shape X_train:(210, 15), Length of y_train:210 +2016-09-08 10:00:06,549 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 10:00:06,549 DEBUG: Info: Shape X_test:(90, 15), Length of y_test:90 +2016-09-08 10:00:06,549 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:06,549 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:06,549 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:06,549 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:06,595 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:06,595 DEBUG: Start: Training +2016-09-08 10:00:06,601 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:06,601 DEBUG: Start: Training +2016-09-08 10:00:06,614 DEBUG: Info: Time for Training: 0.0678429603577[s] +2016-09-08 10:00:06,614 DEBUG: Done: Training +2016-09-08 10:00:06,615 DEBUG: Start: Predicting +2016-09-08 10:00:06,620 DEBUG: Info: Time for Training: 0.0734059810638[s] +2016-09-08 10:00:06,620 DEBUG: Done: Training +2016-09-08 10:00:06,620 DEBUG: Start: Predicting +2016-09-08 10:00:06,621 DEBUG: Done: Predicting +2016-09-08 10:00:06,621 DEBUG: Start: Getting Results +2016-09-08 10:00:06,624 DEBUG: Done: Predicting +2016-09-08 10:00:06,624 DEBUG: Start: Getting Results +2016-09-08 10:00:06,657 DEBUG: Done: Getting Results +2016-09-08 10:00:06,657 DEBUG: Done: Getting Results +2016-09-08 10:00:06,657 INFO: Classification on Fake database for View1 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.5 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.64 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.64 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.5 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.1 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.8 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4625 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.5 + + + Classification took 0:00:00 +2016-09-08 10:00:06,657 INFO: Classification on Fake database for View1 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0714285714286 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0714285714286 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.119522860933 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.04 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.47 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 +2016-09-08 10:00:06,657 INFO: Done: Result Analysis +2016-09-08 10:00:06,657 INFO: Done: Result Analysis +2016-09-08 10:00:06,793 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:06,793 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 10:00:06,793 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:06,794 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:06,794 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 10:00:06,794 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:06,794 DEBUG: Info: Shape X_train:(210, 18), Length of y_train:210 +2016-09-08 10:00:06,794 DEBUG: Info: Shape X_test:(90, 18), Length of y_test:90 +2016-09-08 10:00:06,794 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:06,794 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:06,795 DEBUG: Info: Shape X_train:(210, 18), Length of y_train:210 +2016-09-08 10:00:06,795 DEBUG: Info: Shape X_test:(90, 18), Length of y_test:90 +2016-09-08 10:00:06,795 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:06,795 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:06,834 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:06,834 DEBUG: Start: Training +2016-09-08 10:00:06,837 DEBUG: Info: Time for Training: 0.0438461303711[s] +2016-09-08 10:00:06,837 DEBUG: Done: Training +2016-09-08 10:00:06,837 DEBUG: Start: Predicting +2016-09-08 10:00:06,839 DEBUG: Done: Predicting +2016-09-08 10:00:06,839 DEBUG: Start: Getting Results +2016-09-08 10:00:06,847 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:06,847 DEBUG: Start: Training +2016-09-08 10:00:06,852 DEBUG: Info: Time for Training: 0.0599958896637[s] +2016-09-08 10:00:06,852 DEBUG: Done: Training +2016-09-08 10:00:06,852 DEBUG: Start: Predicting +2016-09-08 10:00:06,855 DEBUG: Done: Predicting +2016-09-08 10:00:06,855 DEBUG: Start: Getting Results +2016-09-08 10:00:06,885 DEBUG: Done: Getting Results +2016-09-08 10:00:06,885 INFO: Classification on Fake database for View2 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.433333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.43956043956 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.43956043956 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.124719695673 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.487804878049 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4375 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + + + Classification took 0:00:00 +2016-09-08 10:00:06,885 INFO: Done: Result Analysis +2016-09-08 10:00:06,896 DEBUG: Done: Getting Results +2016-09-08 10:00:06,896 INFO: Classification on Fake database for View2 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.433333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 9, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.43956043956 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.43956043956 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.124719695673 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.487804878049 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4375 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + + + Classification took 0:00:00 +2016-09-08 10:00:06,897 INFO: Done: Result Analysis +2016-09-08 10:00:07,034 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:07,034 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:07,034 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 10:00:07,034 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 10:00:07,035 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:07,035 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:07,035 DEBUG: Info: Shape X_train:(210, 18), Length of y_train:210 +2016-09-08 10:00:07,035 DEBUG: Info: Shape X_train:(210, 18), Length of y_train:210 +2016-09-08 10:00:07,035 DEBUG: Info: Shape X_test:(90, 18), Length of y_test:90 +2016-09-08 10:00:07,035 DEBUG: Info: Shape X_test:(90, 18), Length of y_test:90 +2016-09-08 10:00:07,035 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:07,035 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:07,036 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:07,036 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:07,067 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:07,067 DEBUG: Start: Training +2016-09-08 10:00:07,067 DEBUG: Info: Time for Training: 0.0337820053101[s] +2016-09-08 10:00:07,067 DEBUG: Done: Training +2016-09-08 10:00:07,068 DEBUG: Start: Predicting +2016-09-08 10:00:07,074 DEBUG: Done: Predicting +2016-09-08 10:00:07,074 DEBUG: Start: Getting Results +2016-09-08 10:00:07,115 DEBUG: Done: Getting Results +2016-09-08 10:00:07,115 INFO: Classification on Fake database for View2 with KNN + +accuracy_score on train : 0.547619047619 +accuracy_score on test : 0.466666666667 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 24 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.547619047619 + - Score on test : 0.466666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.649446494465 + - Score on test : 0.586206896552 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.649446494465 + - Score on test : 0.586206896552 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.452380952381 + - Score on test : 0.533333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.547619047619 + - Score on test : 0.466666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0476928413215 + - Score on test : -0.134839972493 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.515151515152 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.752136752137 + - Score on test : 0.68 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.521229666391 + - Score on test : 0.44 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.452380952381 + - Score on test : 0.533333333333 + + + Classification took 0:00:00 +2016-09-08 10:00:07,115 INFO: Done: Result Analysis +2016-09-08 10:00:07,444 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:07,444 DEBUG: Start: Training +2016-09-08 10:00:07,507 DEBUG: Info: Time for Training: 0.473404169083[s] +2016-09-08 10:00:07,507 DEBUG: Done: Training +2016-09-08 10:00:07,507 DEBUG: Start: Predicting +2016-09-08 10:00:07,514 DEBUG: Done: Predicting +2016-09-08 10:00:07,514 DEBUG: Start: Getting Results +2016-09-08 10:00:07,544 DEBUG: Done: Getting Results +2016-09-08 10:00:07,544 INFO: Classification on Fake database for View2 with RandomForest + +accuracy_score on train : 0.995238095238 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 24, max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.995708154506 + - Score on test : 0.468085106383 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.995708154506 + - Score on test : 0.468085106383 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.990406794809 + - Score on test : -0.109345881217 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.991452991453 + - Score on test : 0.44 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.995726495726 + - Score on test : 0.445 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 +2016-09-08 10:00:07,544 INFO: Done: Result Analysis +2016-09-08 10:00:07,685 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:07,685 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:07,685 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 10:00:07,685 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 10:00:07,685 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:07,686 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:07,686 DEBUG: Info: Shape X_train:(210, 18), Length of y_train:210 +2016-09-08 10:00:07,686 DEBUG: Info: Shape X_train:(210, 18), Length of y_train:210 +2016-09-08 10:00:07,686 DEBUG: Info: Shape X_test:(90, 18), Length of y_test:90 +2016-09-08 10:00:07,686 DEBUG: Info: Shape X_test:(90, 18), Length of y_test:90 +2016-09-08 10:00:07,686 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:07,686 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:07,686 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:07,687 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:07,730 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:07,730 DEBUG: Start: Training +2016-09-08 10:00:07,731 DEBUG: Info: Time for Training: 0.0463371276855[s] +2016-09-08 10:00:07,731 DEBUG: Done: Training +2016-09-08 10:00:07,731 DEBUG: Start: Predicting +2016-09-08 10:00:07,737 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:07,737 DEBUG: Start: Training +2016-09-08 10:00:07,744 DEBUG: Done: Predicting +2016-09-08 10:00:07,744 DEBUG: Start: Getting Results +2016-09-08 10:00:07,758 DEBUG: Info: Time for Training: 0.0734429359436[s] +2016-09-08 10:00:07,758 DEBUG: Done: Training +2016-09-08 10:00:07,758 DEBUG: Start: Predicting +2016-09-08 10:00:07,762 DEBUG: Done: Predicting +2016-09-08 10:00:07,762 DEBUG: Start: Getting Results +2016-09-08 10:00:07,767 DEBUG: Done: Getting Results +2016-09-08 10:00:07,768 INFO: Classification on Fake database for View2 with SGD + +accuracy_score on train : 0.619047619048 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : modified_huber, penalty : l2 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.619047619048 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.701492537313 + - Score on test : 0.592592592593 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.701492537313 + - Score on test : 0.592592592593 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.380952380952 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.619047619048 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.210547659218 + - Score on test : -0.0103806849817 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.622516556291 + - Score on test : 0.551724137931 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.803418803419 + - Score on test : 0.64 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.595257788806 + - Score on test : 0.495 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.380952380952 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 10:00:07,768 INFO: Done: Result Analysis +2016-09-08 10:00:07,797 DEBUG: Done: Getting Results +2016-09-08 10:00:07,797 INFO: Classification on Fake database for View2 with SVMLinear + +accuracy_score on train : 0.47619047619 +accuracy_score on test : 0.555555555556 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.47619047619 + - Score on test : 0.555555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.521739130435 + - Score on test : 0.565217391304 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.521739130435 + - Score on test : 0.565217391304 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.52380952381 + - Score on test : 0.444444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.47619047619 + - Score on test : 0.555555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0568633060564 + - Score on test : 0.119522860933 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.530973451327 + - Score on test : 0.619047619048 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.512820512821 + - Score on test : 0.52 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.471464019851 + - Score on test : 0.56 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.52380952381 + - Score on test : 0.444444444444 + + + Classification took 0:00:00 +2016-09-08 10:00:07,797 INFO: Done: Result Analysis +2016-09-08 10:00:07,935 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:07,935 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMPoly +2016-09-08 10:00:07,935 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:07,935 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:07,935 DEBUG: ### Classification - Database:Fake Feature:View2 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMRBF +2016-09-08 10:00:07,935 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:07,936 DEBUG: Info: Shape X_train:(210, 18), Length of y_train:210 +2016-09-08 10:00:07,936 DEBUG: Info: Shape X_test:(90, 18), Length of y_test:90 +2016-09-08 10:00:07,936 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:07,936 DEBUG: Info: Shape X_train:(210, 18), Length of y_train:210 +2016-09-08 10:00:07,936 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:07,936 DEBUG: Info: Shape X_test:(90, 18), Length of y_test:90 +2016-09-08 10:00:07,936 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:07,937 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:07,984 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:07,985 DEBUG: Start: Training +2016-09-08 10:00:07,990 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:07,990 DEBUG: Start: Training +2016-09-08 10:00:08,003 DEBUG: Info: Time for Training: 0.0686159133911[s] +2016-09-08 10:00:08,003 DEBUG: Done: Training +2016-09-08 10:00:08,003 DEBUG: Start: Predicting +2016-09-08 10:00:08,007 DEBUG: Info: Time for Training: 0.0732269287109[s] +2016-09-08 10:00:08,007 DEBUG: Done: Training +2016-09-08 10:00:08,008 DEBUG: Start: Predicting +2016-09-08 10:00:08,009 DEBUG: Done: Predicting +2016-09-08 10:00:08,009 DEBUG: Start: Getting Results +2016-09-08 10:00:08,012 DEBUG: Done: Predicting +2016-09-08 10:00:08,012 DEBUG: Start: Getting Results +2016-09-08 10:00:08,043 DEBUG: Done: Getting Results +2016-09-08 10:00:08,043 INFO: Classification on Fake database for View2 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.107142857143 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.107142857143 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0298807152334 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.06 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4925 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 +2016-09-08 10:00:08,044 INFO: Done: Result Analysis +2016-09-08 10:00:08,044 DEBUG: Done: Getting Results +2016-09-08 10:00:08,044 INFO: Classification on Fake database for View2 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.687022900763 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.687022900763 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 4.13755692208e-17 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.555555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.9 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 +2016-09-08 10:00:08,044 INFO: Done: Result Analysis +2016-09-08 10:00:08,183 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:08,183 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:08,183 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : Adaboost +2016-09-08 10:00:08,183 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:08,183 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : DecisionTree +2016-09-08 10:00:08,183 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:08,184 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:08,184 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:08,184 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:08,184 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:08,184 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:08,184 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:08,185 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:08,185 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:08,221 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:08,221 DEBUG: Start: Training +2016-09-08 10:00:08,223 DEBUG: Info: Time for Training: 0.0409498214722[s] +2016-09-08 10:00:08,223 DEBUG: Done: Training +2016-09-08 10:00:08,223 DEBUG: Start: Predicting +2016-09-08 10:00:08,226 DEBUG: Done: Predicting +2016-09-08 10:00:08,226 DEBUG: Start: Getting Results +2016-09-08 10:00:08,235 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:08,235 DEBUG: Start: Training +2016-09-08 10:00:08,239 DEBUG: Info: Time for Training: 0.0573840141296[s] +2016-09-08 10:00:08,240 DEBUG: Done: Training +2016-09-08 10:00:08,240 DEBUG: Start: Predicting +2016-09-08 10:00:08,242 DEBUG: Done: Predicting +2016-09-08 10:00:08,243 DEBUG: Start: Getting Results +2016-09-08 10:00:08,267 DEBUG: Done: Getting Results +2016-09-08 10:00:08,268 INFO: Classification on Fake database for View3 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.626262626263 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.626262626263 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.169618786115 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.632653061224 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.585 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 +2016-09-08 10:00:08,268 INFO: Done: Result Analysis +2016-09-08 10:00:08,280 DEBUG: Done: Getting Results +2016-09-08 10:00:08,280 INFO: Classification on Fake database for View3 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.577777777778 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 9, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.577777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.422222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.577777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.145 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5725 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.422222222222 + + + Classification took 0:00:00 +2016-09-08 10:00:08,281 INFO: Done: Result Analysis +2016-09-08 10:00:08,439 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:08,439 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : KNN +2016-09-08 10:00:08,439 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:08,439 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:08,439 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : RandomForest +2016-09-08 10:00:08,440 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:08,440 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:08,440 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:08,440 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:08,441 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:08,441 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:08,441 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:08,441 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:08,441 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:08,489 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:08,489 DEBUG: Start: Training +2016-09-08 10:00:08,490 DEBUG: Info: Time for Training: 0.0522999763489[s] +2016-09-08 10:00:08,490 DEBUG: Done: Training +2016-09-08 10:00:08,490 DEBUG: Start: Predicting +2016-09-08 10:00:08,497 DEBUG: Done: Predicting +2016-09-08 10:00:08,497 DEBUG: Start: Getting Results +2016-09-08 10:00:08,532 DEBUG: Done: Getting Results +2016-09-08 10:00:08,532 INFO: Classification on Fake database for View3 with KNN + +accuracy_score on train : 0.6 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 24 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.6 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.688888888889 + - Score on test : 0.637168141593 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.688888888889 + - Score on test : 0.637168141593 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.4 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.6 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.167225897665 + - Score on test : 0.0487950036474 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.607843137255 + - Score on test : 0.571428571429 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.794871794872 + - Score on test : 0.72 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.574855252275 + - Score on test : 0.5225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.4 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 +2016-09-08 10:00:08,532 INFO: Done: Result Analysis +2016-09-08 10:00:08,857 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:08,857 DEBUG: Start: Training +2016-09-08 10:00:08,918 DEBUG: Info: Time for Training: 0.479932069778[s] +2016-09-08 10:00:08,918 DEBUG: Done: Training +2016-09-08 10:00:08,918 DEBUG: Start: Predicting +2016-09-08 10:00:08,925 DEBUG: Done: Predicting +2016-09-08 10:00:08,925 DEBUG: Start: Getting Results +2016-09-08 10:00:08,961 DEBUG: Done: Getting Results +2016-09-08 10:00:08,961 INFO: Classification on Fake database for View3 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 24, max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.117218854031 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.509090909091 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.56 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4425 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 +2016-09-08 10:00:08,961 INFO: Done: Result Analysis +2016-09-08 10:00:09,084 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:09,084 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SGD +2016-09-08 10:00:09,084 DEBUG: ### Main Programm for Classification MonoView +2016-09-08 10:00:09,084 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:09,085 DEBUG: ### Classification - Database:Fake Feature:View3 train_size:0.7, CrossValidation k-folds:5, cores:1, algorithm : SVMLinear +2016-09-08 10:00:09,085 DEBUG: Start: Determine Train/Test split +2016-09-08 10:00:09,085 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:09,086 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:09,086 DEBUG: Info: Shape X_train:(210, 12), Length of y_train:210 +2016-09-08 10:00:09,086 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:09,086 DEBUG: Info: Shape X_test:(90, 12), Length of y_test:90 +2016-09-08 10:00:09,086 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:09,086 DEBUG: Done: Determine Train/Test split +2016-09-08 10:00:09,086 DEBUG: Start: RandomSearch best settings with 1 iterations +2016-09-08 10:00:09,129 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:09,129 DEBUG: Start: Training +2016-09-08 10:00:09,130 DEBUG: Info: Time for Training: 0.0473730564117[s] +2016-09-08 10:00:09,130 DEBUG: Done: Training +2016-09-08 10:00:09,130 DEBUG: Start: Predicting +2016-09-08 10:00:09,136 DEBUG: Done: RandomSearch best settings +2016-09-08 10:00:09,136 DEBUG: Start: Training +2016-09-08 10:00:09,148 DEBUG: Done: Predicting +2016-09-08 10:00:09,148 DEBUG: Start: Getting Results +2016-09-08 10:00:09,155 DEBUG: Info: Time for Training: 0.0714659690857[s] +2016-09-08 10:00:09,155 DEBUG: Done: Training +2016-09-08 10:00:09,155 DEBUG: Start: Predicting +2016-09-08 10:00:09,158 DEBUG: Done: Predicting +2016-09-08 10:00:09,158 DEBUG: Start: Getting Results +2016-09-08 10:00:09,171 DEBUG: Done: Getting Results +2016-09-08 10:00:09,171 INFO: Classification on Fake database for View3 with SGD + +accuracy_score on train : 0.642857142857 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : modified_huber, penalty : l2 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.642857142857 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.701195219124 + - Score on test : 0.596153846154 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.701195219124 + - Score on test : 0.596153846154 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.357142857143 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.642857142857 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.266179454365 + - Score on test : 0.0456435464588 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.65671641791 + - Score on test : 0.574074074074 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.752136752137 + - Score on test : 0.62 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.628756548111 + - Score on test : 0.5225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.357142857143 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 +2016-09-08 10:00:09,171 INFO: Done: Result Analysis +2016-09-08 10:00:09,189 DEBUG: Done: Getting Results +2016-09-08 10:00:09,190 INFO: Classification on Fake database for View3 with SVMLinear + +accuracy_score on train : 0.47619047619 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.47619047619 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.541666666667 + - Score on test : 0.576923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.541666666667 + - Score on test : 0.576923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.52380952381 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.47619047619 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.068670723144 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.528455284553 + - Score on test : 0.555555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.555555555556 + - Score on test : 0.6 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.465949820789 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.52380952381 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 +2016-09-08 10:00:09,190 INFO: Done: Result Analysis +2016-09-08 10:00:09,488 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:09,488 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:09,489 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:09,490 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:09,491 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:09,492 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:09,493 INFO: Done: Read Database Files +2016-09-08 10:00:09,493 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:09,496 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:09,497 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:09,498 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:09,498 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:09,499 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:09,500 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:09,500 INFO: Done: Read Database Files +2016-09-08 10:00:09,500 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:09,502 INFO: Done: Determine validation split +2016-09-08 10:00:09,502 INFO: Start: Determine 5 folds +2016-09-08 10:00:09,507 INFO: Done: Determine validation split +2016-09-08 10:00:09,508 INFO: Start: Determine 5 folds +2016-09-08 10:00:09,515 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:09,515 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:09,516 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:09,516 INFO: Done: Determine folds +2016-09-08 10:00:09,516 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:09,516 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:09,516 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:09,519 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:09,519 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:09,519 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:09,519 INFO: Done: Determine folds +2016-09-08 10:00:09,519 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:09,520 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:09,520 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:09,608 DEBUG: Done: Random search for SGD +2016-09-08 10:00:09,608 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:09,610 DEBUG: Done: Random search for SGD +2016-09-08 10:00:09,610 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:09,693 DEBUG: Done: Random search for SGD +2016-09-08 10:00:09,693 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:09,696 DEBUG: Done: Random search for SGD +2016-09-08 10:00:09,696 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:09,780 DEBUG: Done: Random search for SGD +2016-09-08 10:00:09,780 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:09,783 DEBUG: Done: Random search for SGD +2016-09-08 10:00:09,783 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:09,866 DEBUG: Done: Random search for SGD +2016-09-08 10:00:09,869 DEBUG: Done: Random search for SGD +2016-09-08 10:00:09,946 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:09,946 INFO: Start: Classification +2016-09-08 10:00:09,947 INFO: Start: Fold number 1 +2016-09-08 10:00:09,974 INFO: Start: Classification +2016-09-08 10:00:09,986 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:09,986 INFO: Start: Classification +2016-09-08 10:00:09,986 INFO: Start: Fold number 1 +2016-09-08 10:00:10,000 INFO: Done: Fold number 1 +2016-09-08 10:00:10,000 INFO: Start: Fold number 2 +2016-09-08 10:00:10,013 INFO: Start: Classification +2016-09-08 10:00:10,026 INFO: Start: Classification +2016-09-08 10:00:10,051 INFO: Done: Fold number 2 +2016-09-08 10:00:10,052 INFO: Start: Fold number 3 +2016-09-08 10:00:10,078 INFO: Start: Classification +2016-09-08 10:00:10,083 INFO: Done: Fold number 1 +2016-09-08 10:00:10,083 INFO: Start: Fold number 2 +2016-09-08 10:00:10,104 INFO: Done: Fold number 3 +2016-09-08 10:00:10,104 INFO: Start: Fold number 4 +2016-09-08 10:00:10,109 INFO: Start: Classification +2016-09-08 10:00:10,130 INFO: Start: Classification +2016-09-08 10:00:10,156 INFO: Done: Fold number 4 +2016-09-08 10:00:10,156 INFO: Start: Fold number 5 +2016-09-08 10:00:10,179 INFO: Done: Fold number 2 +2016-09-08 10:00:10,179 INFO: Start: Fold number 3 +2016-09-08 10:00:10,182 INFO: Start: Classification +2016-09-08 10:00:10,206 INFO: Start: Classification +2016-09-08 10:00:10,208 INFO: Done: Fold number 5 +2016-09-08 10:00:10,208 INFO: Done: Classification +2016-09-08 10:00:10,208 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:10,208 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:10,276 INFO: Done: Fold number 3 +2016-09-08 10:00:10,276 INFO: Start: Fold number 4 +2016-09-08 10:00:10,304 INFO: Start: Classification +2016-09-08 10:00:10,339 INFO: Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 59.6470588235 + -On Test : 50.7317073171 + -On Validation : 47.191011236 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Bayesian Inference using a weight for each view : 0.395042964582, 0.135468886361, 0.187401197987, 0.282086951071 + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + - SGDClassifier with loss : modified_huber, penalty : l2 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:10,339 INFO: Done: Result Analysis +2016-09-08 10:00:10,374 INFO: Done: Fold number 4 +2016-09-08 10:00:10,374 INFO: Start: Fold number 5 +2016-09-08 10:00:10,399 INFO: Start: Classification +2016-09-08 10:00:10,465 INFO: Done: Fold number 5 +2016-09-08 10:00:10,465 INFO: Done: Classification +2016-09-08 10:00:10,466 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:10,466 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:10,588 INFO: Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 55.2941176471 + -On Test : 56.0975609756 + -On Validation : 56.1797752809 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Majority Voting + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + - SGDClassifier with loss : modified_huber, penalty : l2 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:03 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:10,588 INFO: Done: Result Analysis +2016-09-08 10:00:10,737 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:10,738 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:10,738 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:10,739 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:10,740 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:10,740 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:10,741 INFO: Done: Read Database Files +2016-09-08 10:00:10,741 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:10,744 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:10,745 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:10,745 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:10,746 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:10,746 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:10,747 INFO: Done: Determine validation split +2016-09-08 10:00:10,747 INFO: Start: Determine 5 folds +2016-09-08 10:00:10,747 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:10,747 INFO: Done: Read Database Files +2016-09-08 10:00:10,747 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:10,752 INFO: Done: Determine validation split +2016-09-08 10:00:10,752 INFO: Start: Determine 5 folds +2016-09-08 10:00:10,754 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:10,754 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:10,754 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:10,755 INFO: Done: Determine folds +2016-09-08 10:00:10,755 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:10,755 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:10,755 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:10,761 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:10,761 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:10,761 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:10,761 INFO: Done: Determine folds +2016-09-08 10:00:10,761 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:10,762 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:10,762 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:10,811 DEBUG: Done: Random search for SGD +2016-09-08 10:00:10,811 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:10,816 DEBUG: Done: Random search for SGD +2016-09-08 10:00:10,816 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:10,862 DEBUG: Done: Random search for SGD +2016-09-08 10:00:10,862 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:10,867 DEBUG: Done: Random search for SGD +2016-09-08 10:00:10,867 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:10,912 DEBUG: Done: Random search for SGD +2016-09-08 10:00:10,913 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:10,918 DEBUG: Done: Random search for SGD +2016-09-08 10:00:10,918 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:10,962 DEBUG: Done: Random search for SGD +2016-09-08 10:00:10,962 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:10,962 INFO: Start: Classification +2016-09-08 10:00:10,963 INFO: Start: Fold number 1 +2016-09-08 10:00:10,968 DEBUG: Done: Random search for SGD +2016-09-08 10:00:11,027 INFO: Start: Classification +2016-09-08 10:00:11,048 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:11,048 INFO: Start: Classification +2016-09-08 10:00:11,048 INFO: Start: Fold number 1 +2016-09-08 10:00:11,057 INFO: Done: Fold number 1 +2016-09-08 10:00:11,057 INFO: Start: Fold number 2 +2016-09-08 10:00:11,074 INFO: Start: Classification +2016-09-08 10:00:11,103 INFO: Start: Classification +2016-09-08 10:00:11,104 INFO: Done: Fold number 1 +2016-09-08 10:00:11,105 INFO: Start: Fold number 2 +2016-09-08 10:00:11,131 INFO: Start: Classification +2016-09-08 10:00:11,133 INFO: Done: Fold number 2 +2016-09-08 10:00:11,133 INFO: Start: Fold number 3 +2016-09-08 10:00:11,162 INFO: Done: Fold number 2 +2016-09-08 10:00:11,162 INFO: Start: Fold number 3 +2016-09-08 10:00:11,180 INFO: Start: Classification +2016-09-08 10:00:11,189 INFO: Start: Classification +2016-09-08 10:00:11,210 INFO: Done: Fold number 3 +2016-09-08 10:00:11,210 INFO: Start: Fold number 4 +2016-09-08 10:00:11,220 INFO: Done: Fold number 3 +2016-09-08 10:00:11,220 INFO: Start: Fold number 4 +2016-09-08 10:00:11,247 INFO: Start: Classification +2016-09-08 10:00:11,255 INFO: Start: Classification +2016-09-08 10:00:11,277 INFO: Done: Fold number 4 +2016-09-08 10:00:11,277 INFO: Start: Fold number 5 +2016-09-08 10:00:11,285 INFO: Done: Fold number 4 +2016-09-08 10:00:11,286 INFO: Start: Fold number 5 +2016-09-08 10:00:11,304 INFO: Start: Classification +2016-09-08 10:00:11,331 INFO: Start: Classification +2016-09-08 10:00:11,335 INFO: Done: Fold number 5 +2016-09-08 10:00:11,335 INFO: Done: Classification +2016-09-08 10:00:11,335 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:11,335 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:11,361 INFO: Done: Fold number 5 +2016-09-08 10:00:11,362 INFO: Done: Classification +2016-09-08 10:00:11,362 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:11,362 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:11,472 INFO: Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 25.6470588235 + -On Test : 23.4146341463 + -On Validation : 27.4157303371 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Weighted linear using a weight for each view : 1.0, 0.342921905986, 0.474381813597, 0.714066510131 + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + - SGDClassifier with loss : modified_huber, penalty : l2 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:11,473 INFO: Done: Result Analysis +2016-09-08 10:00:11,495 INFO: Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 60.4705882353 + -On Test : 47.8048780488 + -On Validation : 53.2584269663 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with SVM for linear + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + - SGDClassifier with loss : modified_huber, penalty : l2 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:11,495 INFO: Done: Result Analysis +2016-09-08 10:00:11,576 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:11,576 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:11,577 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:11,577 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:11,578 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:11,578 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:11,578 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:11,578 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:11,579 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:11,579 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:11,580 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:11,580 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:11,580 INFO: Done: Read Database Files +2016-09-08 10:00:11,580 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:11,580 INFO: Done: Read Database Files +2016-09-08 10:00:11,580 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:11,586 INFO: Done: Determine validation split +2016-09-08 10:00:11,586 INFO: Done: Determine validation split +2016-09-08 10:00:11,586 INFO: Start: Determine 5 folds +2016-09-08 10:00:11,586 INFO: Start: Determine 5 folds +2016-09-08 10:00:11,596 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:11,596 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:11,596 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:11,596 INFO: Done: Determine folds +2016-09-08 10:00:11,596 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:11,596 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:11,597 DEBUG: Start: Random search for Adaboost with 1 iterations +2016-09-08 10:00:11,597 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:11,597 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:11,598 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:11,598 INFO: Done: Determine folds +2016-09-08 10:00:11,598 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:11,598 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:11,598 DEBUG: Start: Random search for DecisionTree with 1 iterations +2016-09-08 10:00:11,664 DEBUG: Done: Random search for DecisionTree +2016-09-08 10:00:11,752 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:11,752 INFO: Start: Classification +2016-09-08 10:00:11,752 INFO: Start: Fold number 1 +2016-09-08 10:00:11,788 INFO: Start: Classification +2016-09-08 10:00:11,830 INFO: Done: Fold number 1 +2016-09-08 10:00:11,831 INFO: Start: Fold number 2 +2016-09-08 10:00:11,864 INFO: Start: Classification +2016-09-08 10:00:11,867 DEBUG: Done: Random search for Adaboost +2016-09-08 10:00:11,906 INFO: Done: Fold number 2 +2016-09-08 10:00:11,906 INFO: Start: Fold number 3 +2016-09-08 10:00:11,940 INFO: Start: Classification +2016-09-08 10:00:11,944 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:11,944 INFO: Start: Classification +2016-09-08 10:00:11,944 INFO: Start: Fold number 1 +2016-09-08 10:00:11,980 INFO: Done: Fold number 3 +2016-09-08 10:00:11,980 INFO: Start: Fold number 4 +2016-09-08 10:00:11,984 INFO: Start: Classification +2016-09-08 10:00:12,015 INFO: Start: Classification +2016-09-08 10:00:12,026 INFO: Done: Fold number 1 +2016-09-08 10:00:12,026 INFO: Start: Fold number 2 +2016-09-08 10:00:12,056 INFO: Done: Fold number 4 +2016-09-08 10:00:12,057 INFO: Start: Fold number 5 +2016-09-08 10:00:12,066 INFO: Start: Classification +2016-09-08 10:00:12,091 INFO: Start: Classification +2016-09-08 10:00:12,108 INFO: Done: Fold number 2 +2016-09-08 10:00:12,108 INFO: Start: Fold number 3 +2016-09-08 10:00:12,133 INFO: Done: Fold number 5 +2016-09-08 10:00:12,133 INFO: Done: Classification +2016-09-08 10:00:12,133 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:12,133 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:12,147 INFO: Start: Classification +2016-09-08 10:00:12,192 INFO: Done: Fold number 3 +2016-09-08 10:00:12,192 INFO: Start: Fold number 4 +2016-09-08 10:00:12,232 INFO: Start: Classification +2016-09-08 10:00:12,263 INFO: Done: Fold number 4 +2016-09-08 10:00:12,263 INFO: Start: Fold number 5 +2016-09-08 10:00:12,307 INFO: Start: Classification +2016-09-08 10:00:12,350 INFO: Done: Fold number 5 +2016-09-08 10:00:12,351 INFO: Done: Classification +2016-09-08 10:00:12,351 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:12,351 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:12,358 INFO: Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 97.4117647059 + -On Test : 47.3170731707 + -On Validation : 50.7865168539 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 0.564923899429, 0.171414234739, 1.0, 0.282773686486 with monoview classifier : + - Decision Tree with max_depth : 8 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:12,358 INFO: Done: Result Analysis +2016-09-08 10:00:12,486 INFO: Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 100.0 + -On Test : 54.6341463415 + -On Validation : 49.2134831461 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 0.444263234099, 1.0, 0.292116326168, 0.822047817174 with monoview classifier : + - Adaboost with num_esimators : 8, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:12,486 INFO: Done: Result Analysis +2016-09-08 10:00:12,629 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:12,629 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:12,629 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:12,630 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:12,630 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:12,630 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:12,630 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:12,631 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:12,631 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:12,631 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:12,631 INFO: Done: Read Database Files +2016-09-08 10:00:12,631 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:12,631 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:12,632 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:12,632 INFO: Done: Read Database Files +2016-09-08 10:00:12,632 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:12,636 INFO: Done: Determine validation split +2016-09-08 10:00:12,636 INFO: Start: Determine 5 folds +2016-09-08 10:00:12,637 INFO: Done: Determine validation split +2016-09-08 10:00:12,637 INFO: Start: Determine 5 folds +2016-09-08 10:00:12,643 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:12,643 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:12,643 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:12,643 INFO: Done: Determine folds +2016-09-08 10:00:12,643 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:12,643 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:12,643 DEBUG: Start: Random search for KNN with 1 iterations +2016-09-08 10:00:12,644 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:12,644 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:12,644 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:12,644 INFO: Done: Determine folds +2016-09-08 10:00:12,644 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:12,644 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:12,644 DEBUG: Start: Random search for RandomForest with 1 iterations +2016-09-08 10:00:12,692 DEBUG: Done: Random search for KNN +2016-09-08 10:00:12,706 DEBUG: Done: Random search for RandomForest +2016-09-08 10:00:12,736 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:12,736 INFO: Start: Classification +2016-09-08 10:00:12,736 INFO: Start: Fold number 1 +2016-09-08 10:00:12,747 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:12,747 INFO: Start: Classification +2016-09-08 10:00:12,747 INFO: Start: Fold number 1 +2016-09-08 10:00:12,754 INFO: Start: Classification +2016-09-08 10:00:12,768 INFO: Start: Classification +2016-09-08 10:00:12,786 INFO: Done: Fold number 1 +2016-09-08 10:00:12,786 INFO: Start: Fold number 2 +2016-09-08 10:00:12,794 INFO: Done: Fold number 1 +2016-09-08 10:00:12,794 INFO: Start: Fold number 2 +2016-09-08 10:00:12,803 INFO: Start: Classification +2016-09-08 10:00:12,814 INFO: Start: Classification +2016-09-08 10:00:12,836 INFO: Done: Fold number 2 +2016-09-08 10:00:12,836 INFO: Start: Fold number 3 +2016-09-08 10:00:12,839 INFO: Done: Fold number 2 +2016-09-08 10:00:12,839 INFO: Start: Fold number 3 +2016-09-08 10:00:12,853 INFO: Start: Classification +2016-09-08 10:00:12,859 INFO: Start: Classification +2016-09-08 10:00:12,884 INFO: Done: Fold number 3 +2016-09-08 10:00:12,884 INFO: Start: Fold number 4 +2016-09-08 10:00:12,886 INFO: Done: Fold number 3 +2016-09-08 10:00:12,886 INFO: Start: Fold number 4 +2016-09-08 10:00:12,903 INFO: Start: Classification +2016-09-08 10:00:12,903 INFO: Start: Classification +2016-09-08 10:00:12,928 INFO: Done: Fold number 4 +2016-09-08 10:00:12,928 INFO: Start: Fold number 5 +2016-09-08 10:00:12,935 INFO: Done: Fold number 4 +2016-09-08 10:00:12,935 INFO: Start: Fold number 5 +2016-09-08 10:00:12,948 INFO: Start: Classification +2016-09-08 10:00:12,952 INFO: Start: Classification +2016-09-08 10:00:12,974 INFO: Done: Fold number 5 +2016-09-08 10:00:12,974 INFO: Done: Classification +2016-09-08 10:00:12,974 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:12,974 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:12,985 INFO: Done: Fold number 5 +2016-09-08 10:00:12,985 INFO: Done: Classification +2016-09-08 10:00:12,985 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:12,985 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:13,124 INFO: Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 81.0588235294 + -On Test : 43.4146341463 + -On Validation : 48.9887640449 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 0.073909136797, 0.326197494021, 1.0, 0.0290483308675 with monoview classifier : + - Random Forest with num_esimators : 1, max_depth : 8 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:13,124 INFO: Done: Result Analysis +2016-09-08 10:00:13,128 INFO: Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 60.4705882353 + -On Test : 54.6341463415 + -On Validation : 51.9101123596 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 1.0, 0.567673336435, 0.401953729602, 0.0761117950819 with monoview classifier : + - K nearest Neighbors with n_neighbors: 40 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:13,128 INFO: Done: Result Analysis +2016-09-08 10:00:13,279 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:13,279 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:13,279 INFO: ### Main Programm for Multiview Classification +2016-09-08 10:00:13,280 INFO: ### Classification - Database : Fake ; Views : Methyl, MiRNA_, RNASeq, Clinic ; Algorithm : Fusion ; Cores : 1 +2016-09-08 10:00:13,280 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:13,280 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:13,280 INFO: Info: Shape of View0 :(300, 12) +2016-09-08 10:00:13,281 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:13,281 INFO: Info: Shape of View1 :(300, 15) +2016-09-08 10:00:13,281 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:13,281 INFO: Info: Shape of View2 :(300, 18) +2016-09-08 10:00:13,282 INFO: Done: Read Database Files +2016-09-08 10:00:13,282 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:13,282 INFO: Info: Shape of View3 :(300, 12) +2016-09-08 10:00:13,282 INFO: Done: Read Database Files +2016-09-08 10:00:13,282 INFO: Start: Determine validation split for ratio 0.7 +2016-09-08 10:00:13,287 INFO: Done: Determine validation split +2016-09-08 10:00:13,287 INFO: Start: Determine 5 folds +2016-09-08 10:00:13,287 INFO: Done: Determine validation split +2016-09-08 10:00:13,287 INFO: Start: Determine 5 folds +2016-09-08 10:00:13,294 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:13,294 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:13,294 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:13,294 INFO: Done: Determine folds +2016-09-08 10:00:13,294 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:13,294 INFO: Info: Length of Learning Sets: 170 +2016-09-08 10:00:13,294 INFO: Info: Length of Testing Sets: 41 +2016-09-08 10:00:13,294 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:13,294 INFO: Info: Length of Validation Set: 89 +2016-09-08 10:00:13,294 DEBUG: Start: Random search for SGD with 1 iterations +2016-09-08 10:00:13,294 INFO: Done: Determine folds +2016-09-08 10:00:13,295 INFO: Start: Learning with Fusion and 5 folds +2016-09-08 10:00:13,295 INFO: Start: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:13,295 DEBUG: Start: Random search for SVMLinear with 1 iterations +2016-09-08 10:00:13,348 DEBUG: Done: Random search for SVMLinear +2016-09-08 10:00:13,351 DEBUG: Done: Random search for SGD +2016-09-08 10:00:13,416 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:13,416 INFO: Start: Classification +2016-09-08 10:00:13,416 INFO: Start: Fold number 1 +2016-09-08 10:00:13,430 INFO: Done: Randomsearching best settings for monoview classifiers +2016-09-08 10:00:13,430 INFO: Start: Classification +2016-09-08 10:00:13,430 INFO: Start: Fold number 1 +2016-09-08 10:00:13,456 INFO: Start: Classification +2016-09-08 10:00:13,456 INFO: Start: Classification +2016-09-08 10:00:13,485 INFO: Done: Fold number 1 +2016-09-08 10:00:13,485 INFO: Start: Fold number 2 +2016-09-08 10:00:13,492 INFO: Done: Fold number 1 +2016-09-08 10:00:13,492 INFO: Start: Fold number 2 +2016-09-08 10:00:13,518 INFO: Start: Classification +2016-09-08 10:00:13,526 INFO: Start: Classification +2016-09-08 10:00:13,555 INFO: Done: Fold number 2 +2016-09-08 10:00:13,555 INFO: Start: Fold number 3 +2016-09-08 10:00:13,556 INFO: Done: Fold number 2 +2016-09-08 10:00:13,556 INFO: Start: Fold number 3 +2016-09-08 10:00:13,581 INFO: Start: Classification +2016-09-08 10:00:13,598 INFO: Start: Classification +2016-09-08 10:00:13,617 INFO: Done: Fold number 3 +2016-09-08 10:00:13,617 INFO: Start: Fold number 4 +2016-09-08 10:00:13,627 INFO: Done: Fold number 3 +2016-09-08 10:00:13,627 INFO: Start: Fold number 4 +2016-09-08 10:00:13,644 INFO: Start: Classification +2016-09-08 10:00:13,668 INFO: Start: Classification +2016-09-08 10:00:13,682 INFO: Done: Fold number 4 +2016-09-08 10:00:13,682 INFO: Start: Fold number 5 +2016-09-08 10:00:13,698 INFO: Done: Fold number 4 +2016-09-08 10:00:13,698 INFO: Start: Fold number 5 +2016-09-08 10:00:13,709 INFO: Start: Classification +2016-09-08 10:00:13,739 INFO: Start: Classification +2016-09-08 10:00:13,747 INFO: Done: Fold number 5 +2016-09-08 10:00:13,747 INFO: Done: Classification +2016-09-08 10:00:13,747 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:13,747 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:13,769 INFO: Done: Fold number 5 +2016-09-08 10:00:13,769 INFO: Done: Classification +2016-09-08 10:00:13,770 INFO: Info: Time for Classification: 0[s] +2016-09-08 10:00:13,770 INFO: Start: Result Analysis for Fusion +2016-09-08 10:00:13,916 INFO: Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 55.2941176471 + -On Test : 56.0975609756 + -On Validation : 56.1797752809 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 1.0, 0.728775264645, 0.482876097673, 0.365130635662 with monoview classifier : + - SGDClassifier with loss : modified_huber, penalty : l1 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:13,916 INFO: Done: Result Analysis +2016-09-08 10:00:13,962 INFO: Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 57.7647058824 + -On Test : 52.1951219512 + -On Validation : 49.2134831461 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 0.073909136797, 0.326197494021, 1.0, 0.0290483308675 with monoview classifier : + - SVM Linear with C : 3073 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + + +2016-09-08 10:00:13,962 INFO: Done: Result Analysis +2016-09-08 10:00:14,116 DEBUG: Start: Deleting 2 temporary datasets for multiprocessing +2016-09-08 10:00:14,116 DEBUG: Start: Deleting datasets for multiprocessing +2016-09-08 10:00:46,421 INFO: Extraction time : 5.03609514236s, Monoview time : 1473343204.39s, Multiview Time : 9.72813415527s diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..c10750d44ceac649cbe0a6d590287c9f5c97238b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View0 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 9, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.576923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.576923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -2.5337258102e-17 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.555555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.6 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..ba0382dbd9355b910467e7c3ea45cf08fcbdcc56 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.56862745098 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.56862745098 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.00503027272866 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.557692307692 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.58 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5025 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..fc54b504882519b574c965ecf4d8eedeef1cc3aa --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with KNN + +accuracy_score on train : 0.580952380952 +accuracy_score on test : 0.6 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 24 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.580952380952 + - Score on test : 0.6 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.661538461538 + - Score on test : 0.678571428571 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.661538461538 + - Score on test : 0.678571428571 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.419047619048 + - Score on test : 0.4 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.580952380952 + - Score on test : 0.6 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.130162282504 + - Score on test : 0.171735516296 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.601398601399 + - Score on test : 0.612903225806 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.735042735043 + - Score on test : 0.76 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.561069754618 + - Score on test : 0.58 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.419047619048 + - Score on test : 0.4 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..b24a64e65016e9e0263c359295f6bef2f1a42a50 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100004Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 24, max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.52427184466 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.52427184466 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.111088444626 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.509433962264 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.54 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.445 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..9d76693a39a94571bc41ea125ba5b364719e7777 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View1 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.555555555556 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 9, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.555555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.655172413793 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.655172413793 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.444444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.555555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0674199862463 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.575757575758 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.76 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.53 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.444444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..867521e73e098d35580ec0f95af5982d09e98991 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.637168141593 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.637168141593 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0487950036474 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.571428571429 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.72 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..589edf906844f420e689c2580b484c971d25c7bb --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with KNN + +accuracy_score on train : 0.566666666667 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 24 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.566666666667 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.659176029963 + - Score on test : 0.542056074766 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.659176029963 + - Score on test : 0.542056074766 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.433333333333 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.566666666667 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0939782359481 + - Score on test : -0.123737644978 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.586666666667 + - Score on test : 0.508771929825 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.752136752137 + - Score on test : 0.58 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.542735042735 + - Score on test : 0.44 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.433333333333 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..1881529c6a507787f96a3c34d3eae90866b6286f --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SGD + +accuracy_score on train : 0.57619047619 +accuracy_score on test : 0.644444444444 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : modified_huber, penalty : l2 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.57619047619 + - Score on test : 0.644444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.661596958175 + - Score on test : 0.724137931034 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.661596958175 + - Score on test : 0.724137931034 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.42380952381 + - Score on test : 0.355555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.57619047619 + - Score on test : 0.644444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.117819078215 + - Score on test : 0.269679944985 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.595890410959 + - Score on test : 0.636363636364 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.74358974359 + - Score on test : 0.84 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.55459057072 + - Score on test : 0.62 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.42380952381 + - Score on test : 0.355555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..cbbd1a0d3ac70db81be73d8fe7ce021f46f832ab --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMLinear + +accuracy_score on train : 0.485714285714 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.609523809524 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.609523809524 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.485714285714 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0643090141189 + - Score on test : 0.0662541348869 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.533333333333 + - Score on test : 0.581818181818 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.615384615385 + - Score on test : 0.64 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.468982630273 + - Score on test : 0.5325 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.514285714286 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..23cf02f75457800ec7e365c2a0e5f522c0a89edf --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0973655073258 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.617647058824 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.42 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5475 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d276635c447b2d4b0302a0207ec8ee2adf70e5cb --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100005Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View0 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.6 + +Database configuration : + - Database name : Fake + - View name : View0 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.6 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.714285714286 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.714285714286 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.4 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.6 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.171377655346 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.592105263158 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.9 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5625 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.4 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..caa7c9681cce2811ed8c3da610ba1810d410339e --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View2 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.433333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 9, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.43956043956 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.43956043956 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.124719695673 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.487804878049 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4375 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..836977b7ee31de2f3257684bc8347ee0eacbd0e2 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.433333333333 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.43956043956 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.43956043956 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.433333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.124719695673 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.487804878049 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4375 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.566666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..4dbb4c25550cab8b428692b93e320234b1a075c9 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.566666666667 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 24, max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.566666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.621359223301 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.621359223301 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.433333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.566666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.116137919381 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.603773584906 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.64 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5575 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.433333333333 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d66ec8a4ffb7a89486e0947c4e3dbb3653ab71b9 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SGD + +accuracy_score on train : 0.566666666667 +accuracy_score on test : 0.555555555556 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : modified_huber, penalty : l2 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.566666666667 + - Score on test : 0.555555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.640316205534 + - Score on test : 0.666666666667 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.640316205534 + - Score on test : 0.666666666667 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.433333333333 + - Score on test : 0.444444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.566666666667 + - Score on test : 0.555555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.104925880155 + - Score on test : 0.0597614304667 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.595588235294 + - Score on test : 0.571428571429 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.692307692308 + - Score on test : 0.8 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.550454921423 + - Score on test : 0.525 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.433333333333 + - Score on test : 0.444444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..3b75a933e277742b98aa3bc5f78a4a226c873602 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMLinear + +accuracy_score on train : 0.490476190476 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.490476190476 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.566801619433 + - Score on test : 0.576923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.566801619433 + - Score on test : 0.576923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.509523809524 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.490476190476 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0479423260647 + - Score on test : -3.54721613428e-17 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.538461538462 + - Score on test : 0.555555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.598290598291 + - Score on test : 0.6 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.476564653984 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.509523809524 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d128db52bc2a7a322700fc4fe91f03eff1d6eb7f --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.422222222222 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0714285714286 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.0714285714286 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.422222222222 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.119522860933 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.333333333333 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.04 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.47 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.577777777778 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d2a21462829ea28a11dcac00d479a150f15505c0 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100006Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View1 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.5 + +Database configuration : + - Database name : Fake + - View name : View1 View shape : (300, 15) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.64 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.64 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.5 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.1 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.8 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4625 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.5 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d9a3f5a3e6c9f9314127026a077a825115efd21d --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with KNN + +accuracy_score on train : 0.547619047619 +accuracy_score on test : 0.466666666667 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 24 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.547619047619 + - Score on test : 0.466666666667 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.649446494465 + - Score on test : 0.586206896552 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.649446494465 + - Score on test : 0.586206896552 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.452380952381 + - Score on test : 0.533333333333 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.547619047619 + - Score on test : 0.466666666667 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.0476928413215 + - Score on test : -0.134839972493 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.571428571429 + - Score on test : 0.515151515152 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.752136752137 + - Score on test : 0.68 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.521229666391 + - Score on test : 0.44 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.452380952381 + - Score on test : 0.533333333333 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..8e7a8c2dac3a63ea72ac1a1d2b1f00e6057b3021 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with RandomForest + +accuracy_score on train : 0.995238095238 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 24, max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.995708154506 + - Score on test : 0.468085106383 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.995708154506 + - Score on test : 0.468085106383 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.995238095238 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.990406794809 + - Score on test : -0.109345881217 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.991452991453 + - Score on test : 0.44 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.995726495726 + - Score on test : 0.445 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0047619047619 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..9ee2ef1d9e95ac7aee25aa51fcbfbce5d49d2e57 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SGD + +accuracy_score on train : 0.619047619048 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : modified_huber, penalty : l2 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.619047619048 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.701492537313 + - Score on test : 0.592592592593 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.701492537313 + - Score on test : 0.592592592593 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.380952380952 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.619047619048 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.210547659218 + - Score on test : -0.0103806849817 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.622516556291 + - Score on test : 0.551724137931 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.803418803419 + - Score on test : 0.64 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.595257788806 + - Score on test : 0.495 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.380952380952 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d16950a5645f14abfffcd24404af4b15fb8acd2c --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100007Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMLinear + +accuracy_score on train : 0.47619047619 +accuracy_score on test : 0.555555555556 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.47619047619 + - Score on test : 0.555555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.521739130435 + - Score on test : 0.565217391304 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.521739130435 + - Score on test : 0.565217391304 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.52380952381 + - Score on test : 0.444444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.47619047619 + - Score on test : 0.555555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.0568633060564 + - Score on test : 0.119522860933 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.530973451327 + - Score on test : 0.619047619048 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.512820512821 + - Score on test : 0.52 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.471464019851 + - Score on test : 0.56 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.52380952381 + - Score on test : 0.444444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..1fadae0e8031c1efee27d914a774f3215f055b8e --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-Adaboost-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,57 @@ +Classification on Fake database for View3 with Adaboost + +accuracy_score on train : 1.0 +accuracy_score on test : 0.577777777778 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Adaboost with num_esimators : 9, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.577777777778 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.422222222222 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.577777777778 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.145 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5725 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.422222222222 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..ade21df13c32fc925324446cc5c0f45faebc650b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-DecisionTree-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with DecisionTree + +accuracy_score on train : 1.0 +accuracy_score on test : 0.588888888889 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Decision Tree with max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588888888889 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.626262626263 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.626262626263 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.411111111111 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.588888888889 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 0.169618786115 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.632653061224 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.62 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.585 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.411111111111 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-KNN-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-KNN-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..ce39d2332c0af66ff80f32cb189f674d2518b6f0 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-KNN-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with KNN + +accuracy_score on train : 0.6 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - K nearest Neighbors with n_neighbors: 24 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.6 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.688888888889 + - Score on test : 0.637168141593 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.688888888889 + - Score on test : 0.637168141593 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.4 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.6 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.167225897665 + - Score on test : 0.0487950036474 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.607843137255 + - Score on test : 0.571428571429 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.794871794872 + - Score on test : 0.72 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.574855252275 + - Score on test : 0.5225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.4 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..a6d23daee024b72d994eff7b8664df8932567014 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-RandomForest-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with RandomForest + +accuracy_score on train : 1.0 +accuracy_score on test : 0.455555555556 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - Random Forest with num_esimators : 24, max_depth : 25 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.533333333333 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.455555555556 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.117218854031 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.509090909091 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.56 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4425 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.544444444444 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..53f3a855a561b6e9c7a74c9197c38666faa2a1be --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-SVMPoly-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMPoly + +accuracy_score on train : 1.0 +accuracy_score on test : 0.444444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.107142857143 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.107142857143 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.444444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : -0.0298807152334 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.06 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.4925 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.555555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..bf98d86d42889b815483d5075d682abc85a87071 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100008Results-SVMRBF-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View2 with SVMRBF + +accuracy_score on train : 1.0 +accuracy_score on test : 0.544444444444 + +Database configuration : + - Database name : Fake + - View name : View2 View shape : (300, 18) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.687022900763 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 1.0 + - Score on test : 0.687022900763 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 1.0 + - Score on test : 0.544444444444 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 1.0 + - Score on test : 4.13755692208e-17 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.555555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.9 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 1.0 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.0 + - Score on test : 0.455555555556 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100009Results-SGD-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100009Results-SGD-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..04d9aee02671ca751ba32e0fde801c90fc433b58 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100009Results-SGD-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with SGD + +accuracy_score on train : 0.642857142857 +accuracy_score on test : 0.533333333333 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SGDClassifier with loss : modified_huber, penalty : l2 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.642857142857 + - Score on test : 0.533333333333 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.701195219124 + - Score on test : 0.596153846154 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.701195219124 + - Score on test : 0.596153846154 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.357142857143 + - Score on test : 0.466666666667 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.642857142857 + - Score on test : 0.533333333333 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : 0.266179454365 + - Score on test : 0.0456435464588 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.65671641791 + - Score on test : 0.574074074074 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.752136752137 + - Score on test : 0.62 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.628756548111 + - Score on test : 0.5225 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.357142857143 + - Score on test : 0.466666666667 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100009Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100009Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d52376ecd580c0741a63b611a9a5411bd55bef66 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100009Results-SVMLinear-Non-Oui-learnRate0.7-Fake.txt @@ -0,0 +1,54 @@ +Classification on Fake database for View3 with SVMLinear + +accuracy_score on train : 0.47619047619 +accuracy_score on test : 0.511111111111 + +Database configuration : + - Database name : Fake + - View name : View3 View shape : (300, 12) + - Learning Rate : 0.7 + - Labels used : Non, Oui + - Number of cross validation folds : 5 + +Classifier configuration : + - SVM Linear with C : 7704 + - Executed on 1 core(s) + - Got configuration using randomized search with 1 iterations + + + For Accuracy score using None as sample_weights (higher is better) : + - Score on train : 0.47619047619 + - Score on test : 0.511111111111 + For F1 score using None as sample_weights, None as labels, 1 as pos_label, micro as average (higher is better) : + - Score on train : 0.541666666667 + - Score on test : 0.576923076923 + For F-beta score using None as sample_weights, None as labels, 1 as pos_label, micro as average, 1.0 as beta (higher is better) : + - Score on train : 0.541666666667 + - Score on test : 0.576923076923 + For Hamming loss using None as classes (lower is better) : + - Score on train : 0.52380952381 + - Score on test : 0.488888888889 + For Jaccard similarity score using None as sample_weights (higher is better) : + - Score on train : 0.47619047619 + - Score on test : 0.511111111111 + For Log loss using None as sample_weights, 1e-15 as eps (lower is better) : + - Score on train : nan + - Score on test : nan + For Matthews correlation coefficient (higher is better) : + - Score on train : -0.068670723144 + - Score on test : 0.0 + For Precision score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.528455284553 + - Score on test : 0.555555555556 + For Recall score using None as sample_weights, None as labels, 1 as pos_label, binary as average (higher is better) : + - Score on train : 0.555555555556 + - Score on test : 0.6 + For ROC AUC score using None as sample_weights, micro as average (higher is better) : + - Score on train : 0.465949820789 + - Score on test : 0.5 + For Zero one loss using None as sample_weights (lower is better) : + - Score on train : 0.52380952381 + - Score on test : 0.488888888889 + + + Classification took 0:00:00 \ No newline at end of file diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100010Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100010Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..31c5952590b71df2f7b9326a2fed03e501181cfa --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100010Results-Fusion-LateFusion-BayesianInference-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,32 @@ + Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 59.6470588235 + -On Test : 50.7317073171 + -On Validation : 47.191011236 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Bayesian Inference using a weight for each view : 0.395042964582, 0.135468886361, 0.187401197987, 0.282086951071 + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + - SGDClassifier with loss : modified_huber, penalty : l2 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100010Results-Fusion-LateFusion-MajorityVoting-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100010Results-Fusion-LateFusion-MajorityVoting-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..4d7b96cd623963b3db009f942a91b9ba9c0ae0ea --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100010Results-Fusion-LateFusion-MajorityVoting-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,32 @@ + Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 55.2941176471 + -On Test : 56.0975609756 + -On Validation : 56.1797752809 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Majority Voting + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + - SGDClassifier with loss : modified_huber, penalty : l2 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:03 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100011Results-Fusion-LateFusion-SVMForLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100011Results-Fusion-LateFusion-SVMForLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..92e2e7eb2640f7c3de43250d3304533afdc9196b --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100011Results-Fusion-LateFusion-SVMForLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,32 @@ + Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 60.4705882353 + -On Test : 47.8048780488 + -On Validation : 53.2584269663 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with SVM for linear + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + - SGDClassifier with loss : modified_huber, penalty : l2 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100011Results-Fusion-LateFusion-WeightedLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100011Results-Fusion-LateFusion-WeightedLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..9035751820a94da279bbd7183290a6cbe0201fb6 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100011Results-Fusion-LateFusion-WeightedLinear-SGD-SGD-SGD-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,32 @@ + Result for Multiview classification with LateFusion + +Average accuracy : + -On Train : 25.6470588235 + -On Test : 23.4146341463 + -On Validation : 27.4157303371 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : LateFusion with Weighted linear using a weight for each view : 1.0, 0.342921905986, 0.474381813597, 0.714066510131 + -With monoview classifiers : + - SGDClassifier with loss : modified_huber, penalty : l1 + - SGDClassifier with loss : log, penalty : l2 + - SGDClassifier with loss : log, penalty : elasticnet + - SGDClassifier with loss : modified_huber, penalty : l2 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100012Results-Fusion-EarlyFusion-WeightedLinear-Adaboost-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100012Results-Fusion-EarlyFusion-WeightedLinear-Adaboost-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..3ebeedf2c579860400c701b722f7662e73fd3d20 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100012Results-Fusion-EarlyFusion-WeightedLinear-Adaboost-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,31 @@ + Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 100.0 + -On Test : 54.6341463415 + -On Validation : 49.2134831461 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 0.444263234099, 1.0, 0.292116326168, 0.822047817174 with monoview classifier : + - Adaboost with num_esimators : 8, base_estimators : DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None, + max_features=None, max_leaf_nodes=None, min_samples_leaf=1, + min_samples_split=2, min_weight_fraction_leaf=0.0, + presort=False, random_state=None, splitter='best') + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:02 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100012Results-Fusion-EarlyFusion-WeightedLinear-DecisionTree-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100012Results-Fusion-EarlyFusion-WeightedLinear-DecisionTree-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..007add077fe4b19b82933921162236df028dcfe6 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100012Results-Fusion-EarlyFusion-WeightedLinear-DecisionTree-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,28 @@ + Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 97.4117647059 + -On Test : 47.3170731707 + -On Validation : 50.7865168539 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 0.564923899429, 0.171414234739, 1.0, 0.282773686486 with monoview classifier : + - Decision Tree with max_depth : 8 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-KNN-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-KNN-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..d10e58096a314c7404284f1d9762253358d9e56a --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-KNN-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,28 @@ + Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 60.4705882353 + -On Test : 54.6341463415 + -On Validation : 51.9101123596 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 1.0, 0.567673336435, 0.401953729602, 0.0761117950819 with monoview classifier : + - K nearest Neighbors with n_neighbors: 40 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-RandomForest-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-RandomForest-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..3109c44cf165ee1ea89651472f5ebd130af9afd5 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-RandomForest-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,28 @@ + Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 81.0588235294 + -On Test : 43.4146341463 + -On Validation : 48.9887640449 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 0.073909136797, 0.326197494021, 1.0, 0.0290483308675 with monoview classifier : + - Random Forest with num_esimators : 1, max_depth : 8 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..e63a7f8163f3366b568c1f02b93096e9fba2ed76 --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-SGD-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,28 @@ + Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 55.2941176471 + -On Test : 56.0975609756 + -On Validation : 56.1797752809 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 1.0, 0.728775264645, 0.482876097673, 0.365130635662 with monoview classifier : + - SGDClassifier with loss : modified_huber, penalty : l1 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-SVMLinear-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt b/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-SVMLinear-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt new file mode 100644 index 0000000000000000000000000000000000000000..6c39250e0e355ae42944f0ee6289bcb576f2e11d --- /dev/null +++ b/Code/MonoMutliViewClassifiers/Results/20160908-100013Results-Fusion-EarlyFusion-WeightedLinear-SVMLinear-Methyl-MiRNA_-RNASeq-Clinic-MiRNA_-RNASeq-Clinic-Methyl-learnRate0.7-Fake.txt @@ -0,0 +1,28 @@ + Result for Multiview classification with EarlyFusion + +Average accuracy : + -On Train : 57.7647058824 + -On Test : 52.1951219512 + -On Validation : 49.2134831461 + +Dataset info : + -Database name : Fake + -Labels : Methyl, MiRNA_, RNASeq, Clinic + -Views : Methyl, MiRNA_, RNASeq, Clinic + -5 folds + +Classification configuration : + -Algorithm used : EarlyFusion with weighted concatenation, using weights : 0.073909136797, 0.326197494021, 1.0, 0.0290483308675 with monoview classifier : + - SVM Linear with C : 3073 + +Computation time on 1 cores : + Database extraction time : 0:00:00 + Learn Prediction + Fold 1 0:00:00 0:00:00 + Fold 2 0:00:00 0:00:00 + Fold 3 0:00:00 0:00:00 + Fold 4 0:00:00 0:00:00 + Fold 5 0:00:00 0:00:00 + Total 0:00:01 0:00:00 + So a total classification time of 0:00:00. + diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-accuracy_score-20160908-100014.png b/Code/MonoMutliViewClassifiers/Results/Fake-accuracy_score-20160908-100014.png new file mode 100644 index 0000000000000000000000000000000000000000..7f3edfe4fbd6b2ae487dcdc9dba912286b8d8b90 Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-accuracy_score-20160908-100014.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-f1_score-20160908-100017.png b/Code/MonoMutliViewClassifiers/Results/Fake-f1_score-20160908-100017.png new file mode 100644 index 0000000000000000000000000000000000000000..d93e30b1e9ae77cdba769d66ce4ef25942806136 Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-f1_score-20160908-100017.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-fbeta_score-20160908-100019.png b/Code/MonoMutliViewClassifiers/Results/Fake-fbeta_score-20160908-100019.png new file mode 100644 index 0000000000000000000000000000000000000000..f19e280a9410df100c58eecd2686897cf2d2e2e3 Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-fbeta_score-20160908-100019.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-hamming_loss-20160908-100022.png b/Code/MonoMutliViewClassifiers/Results/Fake-hamming_loss-20160908-100022.png new file mode 100644 index 0000000000000000000000000000000000000000..249481bd7ea44e8ce5f98b7f5cfeb41cc7404d1e Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-hamming_loss-20160908-100022.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-jaccard_similarity_score-20160908-100026.png b/Code/MonoMutliViewClassifiers/Results/Fake-jaccard_similarity_score-20160908-100026.png new file mode 100644 index 0000000000000000000000000000000000000000..4f1465baa16805fd5337aad14f75cd3c04a0cfc5 Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-jaccard_similarity_score-20160908-100026.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-log_loss-20160908-100029.png b/Code/MonoMutliViewClassifiers/Results/Fake-log_loss-20160908-100029.png new file mode 100644 index 0000000000000000000000000000000000000000..232e6787f51e89a354d90308ada57cfe2b671100 Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-log_loss-20160908-100029.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-matthews_corrcoef-20160908-100031.png b/Code/MonoMutliViewClassifiers/Results/Fake-matthews_corrcoef-20160908-100031.png new file mode 100644 index 0000000000000000000000000000000000000000..fd927a423dfdf3be17d92144f940a13da1e7b260 Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-matthews_corrcoef-20160908-100031.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-precision_score-20160908-100034.png b/Code/MonoMutliViewClassifiers/Results/Fake-precision_score-20160908-100034.png new file mode 100644 index 0000000000000000000000000000000000000000..3a56169ebbc54882c2118f01b4eb1b3dd8cfa6da Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-precision_score-20160908-100034.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-recall_score-20160908-100037.png b/Code/MonoMutliViewClassifiers/Results/Fake-recall_score-20160908-100037.png new file mode 100644 index 0000000000000000000000000000000000000000..411634514cd1e5248ea7973287b4b15edb3cb1bf Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-recall_score-20160908-100037.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-roc_auc_score-20160908-100040.png b/Code/MonoMutliViewClassifiers/Results/Fake-roc_auc_score-20160908-100040.png new file mode 100644 index 0000000000000000000000000000000000000000..2c537de54fb1cc489785394aef906011f1fa86fc Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-roc_auc_score-20160908-100040.png differ diff --git a/Code/MonoMutliViewClassifiers/Results/Fake-zero_one_loss-20160908-100043.png b/Code/MonoMutliViewClassifiers/Results/Fake-zero_one_loss-20160908-100043.png new file mode 100644 index 0000000000000000000000000000000000000000..8378bf5f9d08d4cf686e321ef86c811adb4da78b Binary files /dev/null and b/Code/MonoMutliViewClassifiers/Results/Fake-zero_one_loss-20160908-100043.png differ