{"id":"W2011376672","doi":"10.1145/1007730.1007737","title":"Class imbalances versus small disjuncts","year":2004,"lang":"en","type":"article","venue":"ACM SIGKDD Explorations Newsletter","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":681,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Computer science; Class (philosophy); Focus (optics); Classifier (UML); Artificial intelligence; Machine learning","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":[0.0092356,0.0008459234,0.001393928,0.001573763,0.001320537,0.003126877,0.001240193,0.00156577,0.003487227],"category_scores_gemma":[0.04035357,0.0003554256,0.0005308785,0.002080924,0.00247638,0.005249341,0.002541616,0.003003834,0.0009469661],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000862524,"about_ca_system_score_gemma":0.0005854605,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003409783,"about_ca_topic_score_gemma":0.0005857431,"domain_scores_codex":[0.9904151,0.002924455,0.0007708457,0.002246215,0.003218947,0.000424373],"domain_scores_gemma":[0.9587988,0.0317678,0.00329871,0.003617362,0.001711249,0.00080604],"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":[0.0030046,0.0007708713,0.03499009,0.0009170283,0.0002916762,0.0005246293,0.0009155563,0.04283617,0.03376814,0.04540804,0.01912977,0.8174434],"study_design_scores_gemma":[0.0004166562,0.002344748,0.03315799,0.0002793549,0.0002936823,0.002319312,0.001582813,0.5031462,0.07975813,0.3345167,0.04200795,0.0001764178],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.368705,0.00465528,0.5936243,0.007358557,0.0009433451,0.0007173375,0.001057946,0.001911701,0.02102649],"genre_scores_gemma":[0.8322751,0.0009214469,0.1607827,0.0008417431,0.0006566821,0.0002880942,0.000735569,0.0002150972,0.003283542],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0092356,"threshold_uncertainty_score":0.04884315,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.068174256664078,"score_gpt":0.279847161916715,"score_spread":0.211672905252637,"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."}}