{"id":"W4220918566","doi":"10.18280/isi.270114","title":"Detecting and Mitigating Bias in Data Using Machine Learning with Pre-Training Metrics","year":2022,"lang":"en","type":"article","venue":"Ingénierie des systèmes d information","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Metric (unit); Divergence (linguistics); Computer science; Performance metric; Sampling bias; Sample (material); Artificial intelligence; Statistics; Pattern recognition (psychology); False positive rate; Work (physics); Machine learning; Sample size determination; Mathematics; Engineering","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001964574,0.0001365537,0.0001686342,0.0006147273,0.0006641185,0.0004488961,0.0007862649,0.0000361016,0.000003673773],"category_scores_gemma":[0.0009281592,0.0001402834,0.00001103397,0.001628475,0.00005701484,0.006311456,0.00114323,0.0003894761,8.269286e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003286038,"about_ca_system_score_gemma":0.0001166054,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002862309,"about_ca_topic_score_gemma":0.00003510884,"domain_scores_codex":[0.9984137,0.0001853318,0.0005021786,0.0002499316,0.0003788304,0.0002699907],"domain_scores_gemma":[0.9986357,0.0002030672,0.0005065513,0.0005296385,0.00008001721,0.00004498845],"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":[0.00003638434,0.00003094787,0.08389889,0.0003513756,0.00003593399,0.0000207213,0.05316053,0.02079126,0.002399838,0.004715634,0.00002179843,0.8345367],"study_design_scores_gemma":[0.000244362,0.00007176583,0.002760379,0.00008651519,0.0000045321,0.0002244187,0.002375305,0.9917577,0.0007947807,0.0005022968,0.0009787768,0.0001991981],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2965907,0.00009561084,0.7023653,0.00003371403,0.00005403553,0.000234501,0.00002635907,0.0003233557,0.0002764371],"genre_scores_gemma":[0.7906161,0.000008364594,0.209119,0.00006748928,0.000009197436,0.0000318961,0.0001354212,0.00000827041,0.000004246851],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9709664,"threshold_uncertainty_score":0.5720593,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08069774931360683,"score_gpt":0.283910502966973,"score_spread":0.2032127536533662,"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."}}