{"id":"W2106909612","doi":"10.5555/777092.777116","title":"Data perturbation for escaping local maxima in learning","year":2002,"lang":"en","type":"article","venue":"","topic":"Bayesian Modeling and Causal Inference","field":"Computer Science","cited_by":81,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Initialization; Machine learning; Computer science; Artificial intelligence; Boosting (machine learning); Maxima; Maxima and minima; Parametric statistics; Test data; Logistic regression; Bayesian probability; Mathematical optimization; Mathematics; Statistics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001998755,0.00005632147,0.00006528806,0.00005032473,0.00005515948,0.00009618913,0.0006247832,0.00003465926,0.00002932822],"category_scores_gemma":[0.00006398872,0.00005190616,0.00001230345,0.0001351211,0.00001176719,0.0005058852,0.0001718041,0.00009536721,0.00003796484],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001767594,"about_ca_system_score_gemma":0.000009537307,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001613873,"about_ca_topic_score_gemma":0.00001704038,"domain_scores_codex":[0.9993198,0.00002434607,0.0001232419,0.0002864053,0.00008612811,0.0001600971],"domain_scores_gemma":[0.9994954,0.00007418917,0.00002276508,0.000350877,0.00002653792,0.00003021815],"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":[0.000002193643,0.00004727962,0.000194544,0.00001528132,0.000003408667,0.000002760503,0.0007266122,0.006528844,0.0002380287,0.1186815,0.002222947,0.8713366],"study_design_scores_gemma":[0.0001301099,0.00002675694,0.00006077659,0.0000134859,7.190602e-7,0.000002732325,0.00003608497,0.9922945,0.00008114974,0.004849977,0.002425642,0.000078087],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.000682177,0.0001238485,0.9939353,0.0008765961,0.00006318429,0.00005822502,7.377669e-7,0.000101117,0.004158776],"genre_scores_gemma":[0.8930503,0.00001755436,0.1053318,0.0002014734,0.0000273174,0.000005980191,0.000006621581,0.000003765899,0.001355145],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9857656,"threshold_uncertainty_score":0.2116672,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1385742530876961,"score_gpt":0.2896610981188911,"score_spread":0.151086845031195,"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."}}