{"id":"W6976652103","doi":"10.60692/zybpg-91c05","title":"Leveraging feature selection to detect potential tax fraudsters","year":2020,"lang":"en","type":"article","venue":"Greater South Information System","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Taxpayer; Treasury; Payment; Enforcement; Evasion (ethics); Agency (philosophy); Feature (linguistics); Work (physics)","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":[0.002502616,0.001427448,0.001505263,0.005652925,0.0007812339,0.001675214,0.001139808,0.001305055,0.0009872796],"category_scores_gemma":[0.006914894,0.0002324017,0.000878519,0.002218243,0.000446725,0.001232579,0.0009235624,0.001046153,0.0007972171],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004200647,"about_ca_system_score_gemma":0.001019963,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003181744,"about_ca_topic_score_gemma":0.002401603,"domain_scores_codex":[0.9984632,0.0003221778,0.0001669173,0.000317793,0.0004483111,0.000281723],"domain_scores_gemma":[0.9954239,0.002458165,0.0006043353,0.0003012994,0.001009211,0.0002030804],"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":[0.001301594,0.001842129,0.1536388,0.0003160504,0.0003527805,0.001850888,0.0003495757,0.03594665,0.01515727,0.001250148,0.01920888,0.7687852],"study_design_scores_gemma":[0.0001090925,0.0004919223,0.04393217,0.0001192611,0.0002971251,0.001219771,0.0004401809,0.930507,0.01201979,0.004122141,0.006672531,0.00006900387],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7747734,0.003528072,0.2073344,0.00257358,0.0005441057,0.000623031,0.002532081,0.003789139,0.004302269],"genre_scores_gemma":[0.9638991,0.000351129,0.03160228,0.0002200049,0.0001919023,0.0001310255,0.002169905,0.00003514005,0.00139943],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005652925,"threshold_uncertainty_score":0.01323527,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02713987545648489,"score_gpt":0.2097561702501647,"score_spread":0.1826162947936798,"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."}}