{"id":"W4412928849","doi":"10.1007/s13385-025-00428-5","title":"From point to probabilistic gradient boosting for claim frequency and severity prediction","year":2025,"lang":"en","type":"article","venue":"European Actuarial Journal","topic":"Probability and Risk Models","field":"Decision Sciences","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université Laval","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Gradient boosting; Boosting (machine learning); Mathematical finance; Probabilistic logic; Artificial intelligence; Point (geometry); Mathematics; Computer science; Pattern recognition (psychology); Econometrics; Machine learning; Random forest; Economics; Financial economics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005548951,0.0008620511,0.001523768,0.001267261,0.0004094092,0.001332097,0.001607063,0.00118222,0.002560954],"category_scores_gemma":[0.01428853,0.0003830007,0.0009969358,0.001303454,0.0009107245,0.001689126,0.001361683,0.001930277,0.001023559],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006245296,"about_ca_system_score_gemma":0.0008988416,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002021817,"about_ca_topic_score_gemma":0.001588695,"domain_scores_codex":[0.998295,0.0008840586,0.00005762763,0.0002358183,0.0004052629,0.0001223296],"domain_scores_gemma":[0.9962908,0.002490367,0.000261887,0.0004607302,0.0003559422,0.0001402577],"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":[0.0004574954,0.00009878336,0.007255525,0.0002220105,0.000188323,0.0001035356,0.00008836745,0.7882024,0.0009188135,0.03816636,0.008372242,0.1559262],"study_design_scores_gemma":[0.00002320636,0.00008033854,0.001008873,0.00003390108,0.00002411545,0.00006603687,0.00001188906,0.9557452,0.0004378986,0.04064295,0.001912399,0.00001325309],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05359148,0.001994183,0.93894,0.0007558226,0.0001708998,0.0001134303,0.0004610652,0.001050164,0.002922816],"genre_scores_gemma":[0.7862583,0.001097378,0.2075256,0.0004393726,0.0003012803,0.0002048742,0.001152963,0.0002061723,0.002814127],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005548951,"threshold_uncertainty_score":0.02934605,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07069668849353687,"score_gpt":0.3447146741924296,"score_spread":0.2740179856988927,"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."}}