{"id":"W3173285506","doi":"10.3390/risks9070126","title":"Improving Explainability of Major Risk Factors in Artificial Neural Networks for Auto Insurance Rate Regulation","year":2021,"lang":"en","type":"article","venue":"Risks","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"","keywords":"Transparency (behavior); Artificial neural network; Variable (mathematics); Computer science; Work (physics); Actuarial science; Insurance industry; Econometrics; Machine learning; Risk analysis (engineering); Artificial intelligence; Economics; Business; Engineering; Mathematics","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":[],"consensus_categories":[],"category_scores_codex":[0.0008842698,0.0001183907,0.0002200402,0.00006734594,0.0001302526,0.00005376336,0.0002851888,0.0001017335,0.000006019128],"category_scores_gemma":[0.001168867,0.0001184456,0.00007789581,0.0004353271,0.00002945558,0.0002544434,0.000125146,0.0003120552,5.205432e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008071741,"about_ca_system_score_gemma":0.00009438303,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002492878,"about_ca_topic_score_gemma":0.0007330936,"domain_scores_codex":[0.9982077,0.0004983033,0.0004342648,0.0004297925,0.0001405483,0.0002894174],"domain_scores_gemma":[0.9983909,0.000564345,0.000292494,0.0005135132,0.0001825194,0.00005630237],"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":[0.00002831267,0.00003828219,0.7182885,0.00009974412,0.000003750328,0.000003078902,0.0007905049,0.173015,0.0004379558,0.001901945,0.000007866835,0.105385],"study_design_scores_gemma":[0.0001062826,0.00002608888,0.4635991,0.000008521983,0.000001562172,5.196145e-7,0.00002192931,0.5337253,0.0009571225,0.001480356,0.00001260347,0.00006059054],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.655058,0.00007762293,0.3439719,0.0002537432,0.0003544106,0.0002191584,0.000007973615,0.00005041423,0.000006690771],"genre_scores_gemma":[0.9917035,0.000003298618,0.008124039,0.00002649277,0.00008083326,0.00002712502,0.00001127866,0.000009957499,0.00001345208],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3607104,"threshold_uncertainty_score":0.483007,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03912109199705017,"score_gpt":0.316868814324062,"score_spread":0.2777477223270118,"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."}}