{"id":"W2150196286","doi":"10.1093/bib/bbr008","title":"Caveats and pitfalls of ROC analysis in clinical microarray research (and how to avoid them)","year":2011,"lang":"en","type":"article","venue":"Briefings in Bioinformatics","topic":"AI in cancer detection","field":"Computer Science","cited_by":134,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Receiver operating characteristic; Ranking (information retrieval); Metric (unit); Gold standard (test); Confidence interval; Statistics; Performance metric; Computer science; Machine learning; Mathematics; Engineering","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":[],"consensus_categories":[],"category_scores_codex":[0.002366798,0.00009893769,0.0002878247,0.0006055884,0.00004882835,0.00009119177,0.0004451882,0.0001131024,0.000003581136],"category_scores_gemma":[0.0002737991,0.00009327553,0.00003872806,0.001680873,0.0001770156,0.000466071,0.0004466394,0.0002895318,0.000006240494],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005691844,"about_ca_system_score_gemma":0.00005577309,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004794655,"about_ca_topic_score_gemma":0.0004067531,"domain_scores_codex":[0.998601,0.00007983766,0.0005277742,0.0002201509,0.0002902686,0.0002810281],"domain_scores_gemma":[0.9989445,0.000270501,0.0001471299,0.0004182032,0.0001229273,0.00009678256],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001850571,0.000363882,0.6453714,0.0006282496,0.0002475155,0.00001915782,0.1041699,0.00008867734,0.0006848801,0.01053704,0.002703226,0.2350011],"study_design_scores_gemma":[0.001585763,0.0008903352,0.874513,0.0003323257,0.00004186556,0.00003961679,0.00138396,0.09952153,0.009044178,0.009913795,0.002124244,0.0006094176],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8420131,0.00006915391,0.1544905,0.001291585,0.00008740427,0.0003433766,0.000003474871,0.00002901515,0.001672405],"genre_scores_gemma":[0.863744,0.0001477123,0.1357143,0.0003139288,0.00001066549,0.00001032792,4.164774e-7,0.000004872081,0.00005376257],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2343916,"threshold_uncertainty_score":0.3803666,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1267116161059898,"score_gpt":0.3407471400724225,"score_spread":0.2140355239664327,"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."}}