{"id":"W4416873823","doi":"10.1109/inc465408.2025.11256487","title":"Fraudulent Firm Classification in External Financial Audits using Machine Learning","year":2025,"lang":"","type":"article","venue":"","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Optech (Canada)","funders":"","keywords":"C4.5 algorithm; Audit; Big data; Interpretability; Naive Bayes classifier; Feature selection; Particle swarm optimization; Scalability; Feature (linguistics)","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":[0.005489764,0.0006659324,0.00103662,0.005622084,0.0005720802,0.002409517,0.0006865544,0.0008149449,0.0003846661],"category_scores_gemma":[0.01741355,0.0002358345,0.0004745443,0.003559113,0.0004622315,0.002095819,0.001000601,0.0008695186,0.000260557],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001012867,"about_ca_system_score_gemma":0.0007937991,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00381631,"about_ca_topic_score_gemma":0.004438718,"domain_scores_codex":[0.9960392,0.001695479,0.0004124044,0.0004111175,0.001060516,0.0003812549],"domain_scores_gemma":[0.9854334,0.007417713,0.00352481,0.001128004,0.002133612,0.0003624397],"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":[0.0005940778,0.0007066447,0.605571,0.000165085,0.0002558444,0.000511771,0.0003374045,0.1204836,0.003005691,0.001325607,0.002034664,0.2650086],"study_design_scores_gemma":[0.00001394719,0.0001709989,0.1209319,0.00007409898,0.00007120636,0.0002651285,0.0004858988,0.8707397,0.003947518,0.002320032,0.0009493809,0.00003013141],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9648297,0.0006153379,0.03189736,0.0005018071,0.00005314476,0.00007950152,0.0003134466,0.0002223182,0.00148738],"genre_scores_gemma":[0.9921041,0.0001076958,0.007347572,0.00002529651,0.00001949186,0.00001117444,0.0002138667,0.00000420633,0.0001664763],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005622084,"threshold_uncertainty_score":0.02903301,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03753821989105183,"score_gpt":0.307735224475259,"score_spread":0.2701970045842071,"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."}}