{"id":"W4367727879","doi":"10.1039/d2dd00146b","title":"Calibration and generalizability of probabilistic models on low-data chemical datasets with DIONYSUS","year":2023,"lang":"en","type":"article","venue":"Digital Discovery","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":48,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Institute for Advanced Research; Vector Institute; University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada; Alliance de recherche numérique du Canada; University of Toronto; Canada Foundation for Innovation; Government of Ontario; Government of Canada; Vector Institute; Canadian Institute for Advanced Research","keywords":"Generalizability theory; Probabilistic logic; Calibration; Computer science; Statistical model; Artificial intelligence; Data mining; Machine learning; Statistics; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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":[],"consensus_categories":[],"category_scores_codex":[0.02139268,0.001608757,0.001732971,0.002677451,0.001101142,0.002663739,0.00364535,0.002415484,0.002787041],"category_scores_gemma":[0.07948272,0.001463164,0.002677954,0.00284794,0.002046233,0.004112083,0.007054157,0.005464204,0.001358045],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001973432,"about_ca_system_score_gemma":0.00331301,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007010524,"about_ca_topic_score_gemma":0.007907437,"domain_scores_codex":[0.9941025,0.003309668,0.0004240906,0.0008931559,0.001107297,0.0001631525],"domain_scores_gemma":[0.9600467,0.03024483,0.001367477,0.006495046,0.001447555,0.0003984086],"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":[0.0008747239,0.0002447628,0.006514367,0.0005446591,0.0007941339,0.0002139009,0.0003172093,0.8057149,0.004984451,0.06627316,0.01202988,0.1014939],"study_design_scores_gemma":[0.00004887053,0.00004249566,0.0004263376,0.00002673691,0.0000245855,0.0000354852,0.00001671283,0.9529029,0.001620651,0.04303635,0.001799043,0.00001986591],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04956146,0.001574759,0.9251287,0.001342513,0.0001386694,0.0002036208,0.002943246,0.01703722,0.002069847],"genre_scores_gemma":[0.3603146,0.001214969,0.6194744,0.001023375,0.0001623423,0.0007025396,0.01136167,0.003158889,0.002587267],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02139268,"threshold_uncertainty_score":0.1131367,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05087454827613958,"score_gpt":0.2897582750390798,"score_spread":0.2388837267629402,"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."}}