{"id":"W3194889096","doi":"10.1002/2050-7038.13046","title":"Gradient boosting and Shapley additive explanations for fraud detection in electricity distribution grids","year":2021,"lang":"en","type":"article","venue":"International Transactions on Electrical Energy Systems","topic":"Electricity Theft Detection Techniques","field":"Engineering","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"International Development Research Centre","funders":"Agência Nacional de Energia Elétrica","keywords":"Boosting (machine learning); Electricity; Scalability; Computer science; Gradient boosting; Feature engineering; Consumption (sociology); Smart grid; Machine learning; Artificial intelligence; Risk analysis (engineering); Environmental economics; Business; Economics; Deep learning; Engineering; Database","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0001731539,0.000211591,0.0002298491,0.000472391,0.0001675057,0.00008918338,0.0001137272,0.0001900547,0.00001468805],"category_scores_gemma":[0.0001663858,0.0002467918,0.0001038387,0.0009388148,0.0000209048,0.0002104423,0.000002792197,0.0003501776,0.000002055622],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009125154,"about_ca_system_score_gemma":0.00003940238,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002421368,"about_ca_topic_score_gemma":0.000374802,"domain_scores_codex":[0.9984894,0.00008454841,0.000444729,0.0003567344,0.0002918461,0.0003327832],"domain_scores_gemma":[0.9989165,0.0005503812,0.00006451516,0.0001236318,0.0002578816,0.0000870849],"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":[0.0004401328,0.0009434074,0.0001296561,0.000118791,0.0007917005,0.00005771441,0.0002213075,0.3092501,0.02486611,0.01978547,0.001114752,0.6422808],"study_design_scores_gemma":[0.0005534973,0.0001959042,0.0003516265,0.00005615818,0.00002394609,0.000129679,0.00002908684,0.8252558,0.1625217,0.0009540363,0.009635241,0.0002933881],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.006346401,0.0003363657,0.9910777,0.0001092701,0.0006903278,0.0002371819,0.0002229568,0.0004698034,0.0005100199],"genre_scores_gemma":[0.997962,0.0005436781,0.0002274758,0.00004382969,0.0001447717,0.0006464223,0.0001587947,0.00003644193,0.0002365642],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9916156,"threshold_uncertainty_score":0.9999985,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008154171754346993,"score_gpt":0.2161172493101611,"score_spread":0.2079630775558141,"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."}}