{"id":"W4414463559","doi":"10.1109/conit65521.2025.11167663","title":"Enhanced Diabetes Diagnosis Using Ensemble Classifiers with Explainable AI and Oversampling for Imbalanced Data","year":2025,"lang":"en","type":"article","venue":"","topic":"Artificial Intelligence in Healthcare","field":"Health Professions","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Random forest; Ensemble learning; Interpretability; Oversampling; Random subspace method; Preprocessor; Feature selection; Classifier (UML); Decision tree","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":["sts"],"consensus_categories":[],"category_scores_codex":[0.0006342622,0.0001904492,0.0003546694,0.0001256252,0.001443863,0.00003255724,0.000295451,0.000207596,0.00009757816],"category_scores_gemma":[0.0008282723,0.0001583423,0.00002433592,0.0003200168,0.0001011247,0.0004071587,0.000307133,0.0003687952,0.000009062567],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002091161,"about_ca_system_score_gemma":0.0005688608,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001661155,"about_ca_topic_score_gemma":0.003239278,"domain_scores_codex":[0.9977838,0.0001394281,0.0005080387,0.0006174555,0.00013851,0.0008127722],"domain_scores_gemma":[0.9962919,0.00237727,0.0001536972,0.0007224178,0.0003206921,0.0001340023],"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":[0.0003776509,0.00008265703,0.9339344,0.002701327,0.000135631,0.000002083768,0.00290248,0.0004135347,0.009165809,0.0171623,0.01151325,0.02160887],"study_design_scores_gemma":[0.00268306,0.0006632928,0.007563623,0.00719796,0.0003920819,7.529391e-7,0.1047459,0.6711301,0.1351609,0.02500643,0.04382974,0.001626215],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7579921,0.0003095192,0.2327796,0.003494194,0.0006442492,0.00228267,0.00008744477,0.0001570408,0.00225319],"genre_scores_gemma":[0.9811561,0.0001151815,0.01293334,0.003788743,0.0001357666,0.0006567859,0.00005597758,0.00003131027,0.001126759],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9263708,"threshold_uncertainty_score":0.9998561,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2270710183652112,"score_gpt":0.4996705948920326,"score_spread":0.2725995765268214,"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."}}