{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001211465,0.001047027,0.0008083389,0.001438023,0.0003701947,0.001195239,0.001602918,0.001589446,0.8309606],"category_scores_gemma":[0.02508954,0.0006921211,0.0009498444,0.001389627,0.0002808754,0.001171964,0.0009872284,0.0009378489,0.2612466],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005800207,"about_ca_system_score_gemma":0.000825256,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003367576,"about_ca_topic_score_gemma":0.007646666,"domain_scores_codex":[0.9996818,0.00008691093,0.00004021306,0.00009631914,0.00006062231,0.0000341955],"domain_scores_gemma":[0.9842726,0.01309129,0.0002782355,0.001058901,0.001113239,0.0001857245],"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":[0.0003921476,0.0001535832,0.002124558,0.002075718,0.00007928132,0.0001066452,0.00004305619,0.006130021,0.000369104,0.001443672,0.9569815,0.03010074],"study_design_scores_gemma":[0.00493095,0.0006413813,0.02049314,0.002751542,0.0003307318,0.0009568847,0.0003384264,0.0906692,0.005527089,0.05463109,0.8184088,0.0003205982],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.0004579541,0.00004579944,0.003628875,0.0001296096,0.00008307686,0.0001027336,0.9921612,0.002043132,0.001347533],"genre_scores_gemma":[0.02006019,0.000208366,0.03234094,0.0006591388,0.0001690214,0.001579395,0.9328629,0.002802573,0.009317504],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8309606,"threshold_uncertainty_score":0.2411141,"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."}}