{"id":"W2909965413","doi":"10.1139/cjce-2018-0117","title":"An extreme gradient boosting method for identifying the factors contributing to crash/near-crash events: a naturalistic driving study","year":2019,"lang":"en","type":"article","venue":"Canadian Journal of Civil Engineering","topic":"Traffic and Road Safety","field":"Engineering","cited_by":58,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Multicollinearity; Crash; Machine learning; Boosting (machine learning); Computer science; Classifier (UML); Artificial intelligence; Statistics; Mathematics; Regression analysis","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00337742,0.0006982893,0.001227797,0.001182238,0.0005264595,0.0005760981,0.0008491842,0.0007183549,0.0005804413],"category_scores_gemma":[0.003235988,0.0002476753,0.0008933167,0.0007064063,0.0003712688,0.0003479439,0.0005488383,0.0007935911,0.000320672],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003310049,"about_ca_system_score_gemma":0.0009121243,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004513358,"about_ca_topic_score_gemma":0.003222691,"domain_scores_codex":[0.9992955,0.0003137038,0.00003337581,0.0001212952,0.000140626,0.00009560711],"domain_scores_gemma":[0.998947,0.0004651849,0.00007405101,0.00009218979,0.0003498214,0.00007172022],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"observational","study_design_scores_codex":[0.000943972,0.0009723824,0.08915384,0.0001868585,0.000423788,0.0006425088,0.0003205158,0.4558002,0.01268636,0.004612618,0.009258715,0.4249983],"study_design_scores_gemma":[0.00001167768,0.00008235835,0.006423373,0.000008825121,0.00002375536,0.000070234,0.0000321122,0.9909184,0.0006889316,0.0008977995,0.0008332889,0.000009254214],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3677248,0.001004194,0.6274316,0.0004981285,0.000160383,0.0002410822,0.0003052205,0.0006283486,0.002006165],"genre_scores_gemma":[0.8895733,0.0003070279,0.1074698,0.0001736863,0.00008427451,0.0001024308,0.0007897271,0.00004381521,0.00145585],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004513358,"threshold_uncertainty_score":0.01786172,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01905352178618235,"score_gpt":0.2513812300012569,"score_spread":0.2323277082150745,"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."}}