{"id":"W3094515489","doi":"10.48550/arxiv.2010.12721","title":"PEP: Parameter Ensembling by Perturbation","year":2020,"lang":"en","type":"preprint","venue":"PubMed","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"National Institute of Biomedical Imaging and Bioengineering","keywords":"Overfitting; MNIST database; Calibration; Computer science; Artificial intelligence; Gaussian; Variance (accounting); Perturbation (astronomy); Mathematics; Set (abstract data type); Statistics; Pattern recognition (psychology); Algorithm; Machine learning; Artificial neural network; Physics","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002930695,0.00202892,0.001164236,0.0009382796,0.0007012485,0.001142033,0.001669924,0.001704291,0.01222338],"category_scores_gemma":[0.01604694,0.000799632,0.0006658251,0.0006002266,0.0008389518,0.002892612,0.001981107,0.003051969,0.00506299],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004841801,"about_ca_system_score_gemma":0.001061249,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001087136,"about_ca_topic_score_gemma":0.001879381,"domain_scores_codex":[0.9983971,0.0006493666,0.00008003363,0.0005136901,0.0002120867,0.0001478066],"domain_scores_gemma":[0.9962355,0.001733979,0.0001300203,0.00132579,0.000444404,0.0001302911],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001376328,0.0005231639,0.007705299,0.0005983272,0.0004130917,0.0005109537,0.0004494244,0.3814786,0.02103899,0.03194198,0.04816595,0.5057979],"study_design_scores_gemma":[0.0000908527,0.0002167744,0.0007945808,0.00008362897,0.00006352638,0.0003022275,0.00005866988,0.9520157,0.0105956,0.02686527,0.008883596,0.00002960901],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06355561,0.001213166,0.8869407,0.0009731308,0.0006370212,0.0004971453,0.0008083861,0.02464794,0.02072684],"genre_scores_gemma":[0.7072449,0.0003564588,0.2766838,0.001051243,0.0002005417,0.0006842575,0.001804218,0.002820979,0.009153523],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01222338,"threshold_uncertainty_score":0.04089135,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05086988193139504,"score_gpt":0.2464434544240062,"score_spread":0.1955735724926112,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}