{"id":"W4390046685","doi":"10.1093/jamia/ocad226","title":"Semi-supervised ROC analysis for reliable and streamlined evaluation of phenotyping algorithms","year":2023,"lang":"en","type":"article","venue":"Journal of the American Medical Informatics Association","topic":"Sepsis Diagnosis and Treatment","field":"Medicine","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada; University of Toronto","keywords":"Computer science; Receiver operating characteristic; Variance (accounting); Software; Artificial intelligence; Machine learning; Data mining; Algorithm","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":[0.04132134,0.002375518,0.002272519,0.006020197,0.0007993871,0.00350975,0.002104588,0.001736857,0.002047627],"category_scores_gemma":[0.1501269,0.0008058348,0.002561682,0.002544644,0.001590498,0.002027988,0.002420352,0.002863818,0.001570486],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001225869,"about_ca_system_score_gemma":0.002934112,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001996219,"about_ca_topic_score_gemma":0.002020851,"domain_scores_codex":[0.9601316,0.02775461,0.002724526,0.004758846,0.004241649,0.0003888163],"domain_scores_gemma":[0.815819,0.1328501,0.01927644,0.01392687,0.01677321,0.001354264],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001359071,0.0006943392,0.09527528,0.002766398,0.003055076,0.0005913536,0.0009791274,0.376243,0.01547436,0.01142453,0.02692367,0.4652138],"study_design_scores_gemma":[0.00007518409,0.0003179756,0.01396492,0.0001966063,0.0001689894,0.0003267359,0.0001062871,0.9578266,0.007413317,0.01372884,0.00575104,0.0001235963],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03358275,0.001233655,0.9567319,0.0004494082,0.00009483092,0.0003792954,0.001800708,0.004639703,0.001087663],"genre_scores_gemma":[0.4254431,0.0005742369,0.5653884,0.0004628539,0.0002552926,0.001321622,0.004840794,0.001139336,0.0005742779],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.04132134,"threshold_uncertainty_score":0.2185307,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06715583229088423,"score_gpt":0.3772528395801668,"score_spread":0.3100970072892825,"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."}}