{"id":"W7107942342","doi":"10.6084/m9.figshare.30741039.v1","title":"Additional file 1 of Augmenting small tabular health data for training prognostic ensemble machine learning models using generative models","year":2025,"lang":"","type":"article","venue":"Open MIND","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University; SickKids Foundation; Ottawa Hospital; Children's Hospital of Eastern Ontario; Hospital for Sick Children; University of Ottawa","funders":"","keywords":"Training set; Training (meteorology); Ensemble learning; Generative grammar; Generative model; Ensemble forecasting","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":["metaepi_narrow","sts","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002735666,0.0005770548,0.001152223,0.0003860395,0.001750171,0.000893884,0.004072286,0.0002253254,0.07050776],"category_scores_gemma":[0.003116675,0.0006645114,0.0001579164,0.001017881,0.0001461466,0.001979633,0.004750758,0.001048063,0.00001343447],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002836119,"about_ca_system_score_gemma":0.006710663,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0009834823,"about_ca_topic_score_gemma":0.0006427761,"domain_scores_codex":[0.9931362,0.001138082,0.001723168,0.002175015,0.0005850833,0.001242405],"domain_scores_gemma":[0.9909195,0.004831077,0.001690962,0.001623977,0.0006241875,0.0003103171],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005514133,0.0001562427,0.00001982728,0.0004370313,0.0001715051,0.000008178891,0.005480122,0.5884184,0.0000263764,0.0007614854,0.02918258,0.3752831],"study_design_scores_gemma":[0.000711116,0.0003463026,0.000007832557,0.004141774,0.00006206946,0.00001524558,0.0007603798,0.9442916,0.00003556261,0.00213975,0.04702537,0.0004629789],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.000411764,0.0008594276,0.729024,0.0008197381,0.0002475559,0.002561566,0.2644077,0.00001123395,0.001657025],"genre_scores_gemma":[0.02228381,0.00001943267,0.8831214,0.0002941567,0.0001813811,0.0003564932,0.09061885,0.00006238398,0.003062125],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.3748201,"threshold_uncertainty_score":0.9995806,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3253474740205635,"score_gpt":0.3868300220797007,"score_spread":0.06148254805913728,"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."}}