{"id":"W4401522263","doi":"10.1080/23270012.2024.2377168","title":"Handling highly imbalanced data for classifying fatality of auto collisions using machine learning techniques","year":2024,"lang":"en","type":"article","venue":"Journal of Management Analytics","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph; Toronto Metropolitan University","funders":"","keywords":"Machine learning; Computer science; Artificial intelligence; Risk analysis (engineering); Data mining; Medicine","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":[0.005259437,0.001020491,0.001464309,0.005129452,0.0009678238,0.002206371,0.001026427,0.0009541007,0.0008996319],"category_scores_gemma":[0.0171287,0.0002861477,0.0009434029,0.003503592,0.000532329,0.002586584,0.00151464,0.001907734,0.0007922012],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008918899,"about_ca_system_score_gemma":0.001397323,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003578493,"about_ca_topic_score_gemma":0.003765176,"domain_scores_codex":[0.9971833,0.0007546097,0.000382189,0.0005503301,0.0008469898,0.0002826359],"domain_scores_gemma":[0.9902774,0.004110355,0.001688331,0.001681214,0.001935375,0.0003072691],"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":[0.0007486009,0.001310377,0.2323079,0.0003214059,0.000405064,0.0005534724,0.0008131056,0.1949964,0.01122348,0.005812597,0.01047023,0.5410373],"study_design_scores_gemma":[0.00002670537,0.0002196516,0.03576578,0.00008542835,0.00007456951,0.0001919249,0.0008056603,0.9336787,0.005619223,0.01895354,0.004520022,0.00005874334],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3791289,0.001505596,0.6073493,0.002173819,0.0006037399,0.0004789835,0.004142576,0.00169259,0.002924414],"genre_scores_gemma":[0.8689948,0.0005570974,0.1241795,0.0002437153,0.0003197488,0.0002278007,0.004565964,0.00007560548,0.0008358176],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005259437,"threshold_uncertainty_score":0.02781492,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09543007381265481,"score_gpt":0.3582703427381831,"score_spread":0.2628402689255284,"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."}}