{"id":"W4321214126","doi":"10.1186/s13040-023-00322-4","title":"The Matthews correlation coefficient (MCC) should replace the ROC AUC as the standard metric for assessing binary classification","year":2023,"lang":"en","type":"article","venue":"BioData Mining","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":513,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Receiver operating characteristic; Matthews correlation coefficient; Binary classification; Statistics; Confusion matrix; False positive rate; Artificial intelligence; Correlation; Mathematics; Binary number; Sensitivity (control systems); Metric (unit); Classifier (UML); Computer science; Pattern recognition (psychology); Machine learning; Support vector machine; Arithmetic","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":["sts","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.004527626,0.0001897326,0.0001500108,0.0002021247,0.002025972,0.001443565,0.002546804,0.0001023849,0.000004457973],"category_scores_gemma":[0.00176413,0.0001014715,0.00007332858,0.002370559,0.0001822049,0.0008307789,0.0005902674,0.0002358752,0.00008865461],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001672805,"about_ca_system_score_gemma":0.0002291622,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001151915,"about_ca_topic_score_gemma":0.000005820631,"domain_scores_codex":[0.9975507,0.0002870343,0.0004909845,0.0006005724,0.0006539651,0.0004167016],"domain_scores_gemma":[0.9942104,0.002775441,0.0004784605,0.002267737,0.0002188182,0.00004914254],"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":[0.00009181811,0.00006502582,0.001617078,0.00003878062,0.00006402737,0.00000426817,0.002719463,0.0005908711,0.009261704,0.157185,0.4464743,0.3818877],"study_design_scores_gemma":[0.0002458281,0.000102889,0.01393474,0.00005681081,0.00002881783,0.0000147217,0.002323982,0.6029679,0.003380282,0.001024668,0.375668,0.0002514689],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.003982812,0.0002232328,0.9644635,0.02789442,0.0008631662,0.001040715,0.0001058182,0.0008031839,0.0006231703],"genre_scores_gemma":[0.9387665,0.0003413442,0.05482959,0.001547972,0.0002815714,0.0009967815,0.0007729403,0.00006939482,0.002393937],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9347836,"threshold_uncertainty_score":0.999593,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1099873591318314,"score_gpt":0.3627967148767833,"score_spread":0.2528093557449519,"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."}}