{"id":"W4414447058","doi":"10.1016/j.infsof.2025.107896","title":"Accuracy-fairness trade-off in ML for healthcare: A quantitative evaluation of bias mitigation strategies","year":2025,"lang":"en","type":"article","venue":"Information and Software Technology","topic":"Ethics and Social Impacts of AI","field":"Social Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"","keywords":"Debiasing; Benchmarking; Boosting (machine learning); Value-Based Purchasing; Health care; Metric (unit); Adversarial system; Selection bias","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.002349364,0.00006465935,0.0001546173,0.0004515962,0.0002457465,0.00007949139,0.0001122599,0.0003295807,0.000004444744],"category_scores_gemma":[0.008456049,0.00006681753,0.00002661754,0.0006227676,0.0003427769,0.001321289,0.00001811059,0.0001475679,7.557113e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001099596,"about_ca_system_score_gemma":0.001058027,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0009328806,"about_ca_topic_score_gemma":0.006222489,"domain_scores_codex":[0.9990287,0.0001309802,0.0003657271,0.00007458673,0.000245273,0.000154776],"domain_scores_gemma":[0.9983888,0.0005910763,0.0002218956,0.00007976499,0.000694943,0.00002355663],"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":[0.0000138873,0.00001167281,0.001952531,0.00006831402,0.000007367854,2.903891e-8,0.02038091,0.00003152709,0.000005219072,0.7411061,0.0000886207,0.2363338],"study_design_scores_gemma":[0.0009419692,0.0001190918,0.01045469,0.000132085,0.00002008077,2.068328e-7,0.1159989,0.001291225,0.000178484,0.8627868,0.007965527,0.0001109504],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8656334,0.001472301,0.0207616,0.1049242,0.0003792745,0.002078158,0.00005003331,0.0003017819,0.004399287],"genre_scores_gemma":[0.9972688,0.0004356344,0.001829618,0.0003672414,0.000006357767,0.00005905115,0.0000239268,0.000002037873,0.000007335467],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2362229,"threshold_uncertainty_score":0.9998962,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07906430964697729,"score_gpt":0.435428972833954,"score_spread":0.3563646631869767,"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."}}