{"id":"W1982069553","doi":"10.5539/cis.v8n1p151","title":"An Empirical Analysis of Imbalanced Data Classification","year":2015,"lang":"en","type":"article","venue":"Computer and Information Science","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Regina","funders":"","keywords":"Computer science; Naive Bayes classifier; Support vector machine; Artificial intelligence; Machine learning; Classifier (UML); Decision tree; Empirical research; Data mining; Pattern recognition (psychology); Statistics; Mathematics","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.001592321,0.00008231695,0.0001633489,0.0008119964,0.00009880779,0.0004105243,0.002700826,0.00003594055,0.000001444674],"category_scores_gemma":[0.00009977828,0.00007185402,0.00001637084,0.003475669,0.0002849379,0.02341259,0.0006613145,0.00005954544,0.000009267014],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004018844,"about_ca_system_score_gemma":0.0002394872,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00000736522,"about_ca_topic_score_gemma":7.558975e-7,"domain_scores_codex":[0.9984975,0.00003998246,0.0004081222,0.0003257464,0.0005734852,0.00015518],"domain_scores_gemma":[0.9971024,0.00003597591,0.0002521249,0.001840648,0.000583965,0.0001848192],"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":[0.00001739222,0.0001727008,0.04424514,0.00002563498,0.00007061092,5.707203e-7,0.00697292,0.00119171,0.002250922,0.1936006,0.007837658,0.7436142],"study_design_scores_gemma":[0.00009132869,0.00004849927,0.1889763,0.000002308873,0.00001074273,0.000002025088,0.00004317251,0.8072237,0.0005689144,0.0001942509,0.00276292,0.00007584034],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01585666,0.000007242279,0.9825162,0.0002870908,0.0001247157,0.00009340716,0.0000300004,0.0001593185,0.0009253287],"genre_scores_gemma":[0.764784,0.00001024247,0.234508,0.0004789567,0.00001361382,0.000003617088,0.0001992398,0.000001049337,0.000001207341],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8060319,"threshold_uncertainty_score":0.9902464,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1120192065287858,"score_gpt":0.3778644614599534,"score_spread":0.2658452549311676,"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."}}