{"id":"W4410406004","doi":"10.1186/s12874-025-02561-x","title":"Comparison of methods for tuning machine learning model hyper-parameters: with application to predicting high-need high-cost health care users","year":2025,"lang":"en","type":"article","venue":"BMC Medical Research Methodology","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute for Work & Health; Institute for Clinical Evaluative Sciences; University of Toronto","funders":"","keywords":"Machine learning; Bayesian optimization; Computer science; Artificial intelligence; Random forest; Gradient boosting; Overfitting; Boosting (machine learning); Extreme learning machine; Cross-validation; Calibration; Statistics; Mathematics; Artificial neural network","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":[0.01366438,0.001313586,0.001101271,0.001320893,0.0004681547,0.001005858,0.001518745,0.001599775,0.0007454448],"category_scores_gemma":[0.02723085,0.0005851276,0.00131395,0.0008658766,0.0005658374,0.001190725,0.001276219,0.002087368,0.000289834],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001156428,"about_ca_system_score_gemma":0.001616869,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005448135,"about_ca_topic_score_gemma":0.0037275,"domain_scores_codex":[0.996402,0.002281951,0.0002184026,0.0005462863,0.0004104026,0.0001408566],"domain_scores_gemma":[0.9859504,0.0107776,0.0008235763,0.0009055965,0.001306225,0.0002365645],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0009905886,0.0004265285,0.01733076,0.0002768651,0.0006641946,0.00005954521,0.0001756026,0.8614542,0.001528453,0.001508525,0.001318962,0.1142657],"study_design_scores_gemma":[0.0001106331,0.0002095767,0.003983952,0.00006717601,0.00008304486,0.00004186492,0.00003636136,0.9913668,0.001744468,0.001871571,0.0004555218,0.0000290694],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.3690068,0.004353611,0.620203,0.001047908,0.0001439397,0.0006149736,0.0003942326,0.001404929,0.002830653],"genre_scores_gemma":[0.7968217,0.0008101901,0.2002774,0.0003448474,0.00005732069,0.0004460087,0.0005262551,0.0002022372,0.0005140306],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01366438,"threshold_uncertainty_score":0.07226503,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3850406573113828,"score_gpt":0.6055089837650878,"score_spread":0.220468326453705,"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."}}