{"id":"W4404716839","doi":"10.1007/s44163-024-00199-0","title":"The effect of resampling techniques on the performances of machine learning clinical risk prediction models in the setting of severe class imbalance: development and internal validation in a retrospective cohort","year":2024,"lang":"en","type":"article","venue":"Discover Artificial Intelligence","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa; Mount Sinai Hospital; University of British Columbia; Carleton University; Providence Health Care; St. Paul's Hospital; University of Toronto; Dalhousie University","funders":"Natural Sciences and Engineering Research Council of Canada; Nova Scotia Health Authority; Dalhousie University; Providence Health Care","keywords":"Resampling; Oversampling; Random forest; Machine learning; Artificial intelligence; Logistic regression; Boosting (machine learning); Computer science; Jackknife resampling; Algorithm; Data mining; Statistics; Mathematics; Bandwidth (computing)","routes":{"ca_aff":true,"ca_fund":true,"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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.06921457,0.001109619,0.0009043601,0.001013253,0.0006446784,0.001056387,0.00112104,0.0008869751,0.0002976214],"category_scores_gemma":[0.1139185,0.0003917957,0.001513425,0.0006338632,0.001130175,0.0007402333,0.001156166,0.001361384,0.0002000665],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005699329,"about_ca_system_score_gemma":0.001080475,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001448449,"about_ca_topic_score_gemma":0.001262651,"domain_scores_codex":[0.9777446,0.01578289,0.00178131,0.002170246,0.002215731,0.0003052127],"domain_scores_gemma":[0.9000435,0.07167244,0.007315835,0.01407562,0.006289853,0.0006027436],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.005963742,0.0008624533,0.7177564,0.0004392921,0.003289392,0.0004309686,0.001369806,0.1124511,0.01300873,0.001119612,0.002255748,0.1410527],"study_design_scores_gemma":[0.000605316,0.006848752,0.3264685,0.0004165391,0.001861172,0.001574818,0.0006105693,0.609082,0.0447432,0.003386129,0.004182344,0.0002205798],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9155381,0.001752545,0.0807226,0.0002636572,0.0001436127,0.0003738062,0.0003898283,0.000365878,0.0004498578],"genre_scores_gemma":[0.9696097,0.0002518881,0.02885789,0.00009453743,0.00004890432,0.0002169883,0.0007858606,0.00005570766,0.00007861578],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9307854,"threshold_uncertainty_score":0.366046,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03163749091959291,"score_gpt":0.32662512621877,"score_spread":0.2949876352991772,"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."}}