{"id":"W4402836358","doi":"10.2139/ssrn.4967592","title":"Oriented Data-Generating Processes: A Categorization of ROC Curves","year":2024,"lang":"en","type":"preprint","venue":"SSRN Electronic Journal","topic":"Statistical and Computational Modeling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université Laval","funders":"","keywords":"Categorization; Computer science; Receiver operating characteristic; Natural language processing; Artificial intelligence; Data mining; Machine learning","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":["research_integrity"],"consensus_categories":[],"category_scores_codex":[0.001237,0.0002164619,0.0002841836,0.0001584223,0.0001176815,0.0002066673,0.001630632,0.00009282787,0.00000368735],"category_scores_gemma":[0.0003677075,0.0001862007,0.00006145555,0.0004887545,0.00002521379,0.0002532053,0.001874166,0.002304577,0.000009465181],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002833152,"about_ca_system_score_gemma":0.01089263,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005228505,"about_ca_topic_score_gemma":0.00009876907,"domain_scores_codex":[0.9972118,0.00008628536,0.0006086951,0.0005808441,0.0005954804,0.0009168522],"domain_scores_gemma":[0.9983674,0.0001199074,0.0003632812,0.000470385,0.000612052,0.00006701273],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000007038195,0.0000677924,0.00001799821,0.0012037,0.0002505415,0.000008046908,0.0002658118,0.03506155,0.00005477184,0.9250655,0.0002629472,0.03773428],"study_design_scores_gemma":[0.00006057839,0.00004535385,0.000001880497,0.0004576666,0.00003920943,0.00008449097,0.00003878796,0.4648269,0.00002293922,0.5342649,0.00003824399,0.0001190666],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001272666,0.0235381,0.9731998,0.0009592457,0.0006850724,0.000139786,0.00003654373,0.0000764807,0.00009233731],"genre_scores_gemma":[0.9240968,0.01055431,0.06355778,0.0002097538,0.0008620225,0.00002146261,0.0003981384,0.00004209522,0.0002576288],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9228241,"threshold_uncertainty_score":0.9999971,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02794513490830918,"score_gpt":0.2940385003390201,"score_spread":0.2660933654307109,"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."}}