{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006528873,0.0001172333,0.0002061588,0.0001223853,0.0001174202,0.0001081497,0.0006044624,0.0000593431,8.541048e-7],"category_scores_gemma":[0.0006199299,0.00006195175,0.00004838058,0.0005300564,0.0001936029,0.0004068573,0.0001541839,0.0005406842,5.573212e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007303135,"about_ca_system_score_gemma":0.00005693386,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001654924,"about_ca_topic_score_gemma":0.0001025189,"domain_scores_codex":[0.9977745,0.0005520465,0.0008506921,0.0003036979,0.0003813116,0.0001377365],"domain_scores_gemma":[0.9976659,0.001580475,0.0003497099,0.0003260103,0.00006699748,0.00001090877],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001916315,0.00007515171,0.6421362,0.0001130936,0.00004643762,0.00000141868,0.007321418,0.005873651,0.001380186,0.08861393,0.00002481024,0.254222],"study_design_scores_gemma":[0.00002136319,0.0003270698,0.07161967,0.0005042452,0.000008545559,0.000002384326,0.0003459596,0.6889395,0.2275269,0.01059556,0.00003202668,0.00007678719],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.64902,0.0001061261,0.3499114,0.0001575065,0.00009162219,0.0004624238,0.0000157788,0.00003693724,0.0001982174],"genre_scores_gemma":[0.9977924,0.0004151034,0.00166748,0.000007981775,0.00002194967,0.00007846684,0.000006528637,0.000005793975,0.000004252764],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6830658,"threshold_uncertainty_score":0.2526319,"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."}}