{"id":"W7107942453","doi":"10.6084/m9.figshare.30741039","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":"Figshare","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":["metaresearch","metaepi_narrow","sts","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.000717604,0.0006609031,0.001049551,0.0004709678,0.001611655,0.000409246,0.002584721,0.0002965216,0.5383527],"category_scores_gemma":[0.01444749,0.0007804869,0.0002434063,0.001163715,0.00003836863,0.001358752,0.002804941,0.0012873,0.00003499379],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003193622,"about_ca_system_score_gemma":0.005589104,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002168902,"about_ca_topic_score_gemma":0.0001811415,"domain_scores_codex":[0.9935108,0.0008690549,0.001561561,0.001967087,0.0006946676,0.001396792],"domain_scores_gemma":[0.9878831,0.007376081,0.001762211,0.001603873,0.001036825,0.0003378973],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001576317,0.00006967196,0.000002733037,0.002562826,0.0001010347,0.000006505954,0.001401482,0.427978,0.000003358253,0.0004100185,0.5296158,0.03783282],"study_design_scores_gemma":[0.0004252739,0.0002363895,0.00001093883,0.02489771,0.00002920179,0.00001312733,0.0001862311,0.8853134,0.00001036159,0.001624055,0.0867742,0.0004791234],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.000005533961,0.002029475,0.2278871,0.0003769963,0.0001328943,0.001574256,0.7674115,0.0001409122,0.0004412821],"genre_scores_gemma":[0.003998619,0.000006928768,0.2494938,0.0005080251,0.0002446531,0.0009939286,0.7439733,0.00006930501,0.0007114743],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.5383177,"threshold_uncertainty_score":0.9996881,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.326985855018233,"score_gpt":0.3529429090825292,"score_spread":0.02595705406429621,"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."}}