From fcab676c7fa5928038ed57c8780e1e4eb2768b29 Mon Sep 17 00:00:00 2001 From: Baptiste Bauvin <baptiste.bauvin@lis-lab.fr> Date: Wed, 25 Jul 2018 09:29:18 -0400 Subject: [PATCH] Modified possible HP for CQB and CQBv2 --- .../MonoMultiViewClassifiers/MonoviewClassifiers/CQBoost.py | 6 +++--- .../MonoviewClassifiers/CQBoostv2.py | 4 ++-- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/multiview_platform/MonoMultiViewClassifiers/MonoviewClassifiers/CQBoost.py b/multiview_platform/MonoMultiViewClassifiers/MonoviewClassifiers/CQBoost.py index b683dd01..b53ab517 100644 --- a/multiview_platform/MonoMultiViewClassifiers/MonoviewClassifiers/CQBoost.py +++ b/multiview_platform/MonoMultiViewClassifiers/MonoviewClassifiers/CQBoost.py @@ -311,8 +311,8 @@ def paramsToSet(nIter, randomState): """Used for weighted linear early fusion to generate random search sets""" paramsSet = [] for _ in range(nIter): - paramsSet.append({"mu": randomState.choice([0.001, 0.002]), - "epsilon": randomState.choice([1e-08, 2e-08]), + paramsSet.append({"mu": randomState.uniform(1e-02, 10**(-0.5)), + "epsilon": 10**-randomState.randint(1, 15), "n_max_iterations": None}) return paramsSet @@ -344,7 +344,7 @@ def genBestParams(detector): def genParamsFromDetector(detector): nIter = len(detector.cv_results_['param_classifier__mu']) - return [("mu", np.array([0.001 for _ in range(nIter)])), + return [("mu", detector.cv_results_['param_classifier__mu']), ("epsilon", np.array(detector.cv_results_['param_classifier__epsilon'])), ("n_max_iterations", np.array(detector.cv_results_['param_classifier__n_max_iterations']))] diff --git a/multiview_platform/MonoMultiViewClassifiers/MonoviewClassifiers/CQBoostv2.py b/multiview_platform/MonoMultiViewClassifiers/MonoviewClassifiers/CQBoostv2.py index d2b2b0d8..0039a4f4 100644 --- a/multiview_platform/MonoMultiViewClassifiers/MonoviewClassifiers/CQBoostv2.py +++ b/multiview_platform/MonoMultiViewClassifiers/MonoviewClassifiers/CQBoostv2.py @@ -330,8 +330,8 @@ def paramsToSet(nIter, randomState): """Used for weighted linear early fusion to generate random search sets""" paramsSet = [] for _ in range(nIter): - paramsSet.append({"mu": randomState.choice([0.001, 0.002]), - "epsilon": randomState.choice([1e-08, 2e-08]), + paramsSet.append({"mu": randomState.uniform(1e-02, 10**(-0.5)), + "epsilon": 10**-randomState.randint(1, 15), "n_max_iterations": None}) return paramsSet -- GitLab