{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001228121,0.0009681074,0.0007243769,0.001268477,0.0003565346,0.001042747,0.001493703,0.001576457,0.7869102],"category_scores_gemma":[0.02610083,0.0006260226,0.0008523822,0.001244817,0.0002605809,0.001044109,0.0009201471,0.0008812606,0.1983217],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000547781,"about_ca_system_score_gemma":0.000889737,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00408958,"about_ca_topic_score_gemma":0.009161499,"domain_scores_codex":[0.9996613,0.00008823544,0.00004714164,0.0001046343,0.00006436821,0.00003424904],"domain_scores_gemma":[0.9830706,0.01407274,0.000348182,0.001105862,0.001217499,0.0001851199],"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":[0.0004443279,0.0002254126,0.003244663,0.002314684,0.00009065195,0.0001299436,0.00005359341,0.006656337,0.0004777703,0.001407194,0.9504767,0.03447874],"study_design_scores_gemma":[0.00564284,0.0007593188,0.028712,0.003088638,0.0003905452,0.001175327,0.000472498,0.1033558,0.006338217,0.05616258,0.7935612,0.0003410638],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.0005382107,0.00003480739,0.003125852,0.000118281,0.00006590087,0.0001032899,0.9938132,0.00122976,0.0009707602],"genre_scores_gemma":[0.02036796,0.0001582511,0.02869292,0.0005503183,0.0001413173,0.001667619,0.9398654,0.001494289,0.007061942],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7869102,"threshold_uncertainty_score":0.3039467,"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."}}