{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0553689,0.001702814,0.001292779,0.001834609,0.001185693,0.003107067,0.00139459,0.002111899,0.001519589],"category_scores_gemma":[0.1741492,0.0003825491,0.0008660177,0.001040132,0.003037669,0.00362214,0.003475047,0.002264847,0.0003701933],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00222056,"about_ca_system_score_gemma":0.002786881,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002105179,"about_ca_topic_score_gemma":0.00183167,"domain_scores_codex":[0.9608954,0.02630044,0.001697471,0.003174879,0.006797224,0.001134516],"domain_scores_gemma":[0.8357735,0.1366554,0.006453228,0.01114085,0.008159628,0.001817313],"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":[0.003588784,0.0007489579,0.08033689,0.001291761,0.001048575,0.0002309593,0.001522417,0.3970768,0.009043346,0.03941482,0.006023636,0.4596731],"study_design_scores_gemma":[0.0002696079,0.003477624,0.02304812,0.0006422736,0.0004403552,0.0004449241,0.0007532472,0.8565961,0.02715441,0.08024055,0.006751935,0.00018083],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3292869,0.008422211,0.6414955,0.006307635,0.0006137959,0.0009182715,0.0006596062,0.001176312,0.01111983],"genre_scores_gemma":[0.9054244,0.0005180351,0.09167718,0.0007740579,0.0001826985,0.0002824551,0.0002505145,0.0001124873,0.000778257],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0553689,"threshold_uncertainty_score":0.2928222,"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."}}