{"id":"W2001536048","doi":"10.1109/icmla.2012.144","title":"A Dynamic Sampling Framework for Multi-class Imbalanced Data","year":2012,"lang":"en","type":"article","venue":"","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"","keywords":"Sampling (signal processing); Computer science; Classifier (UML); Data mining; Machine learning; Oversampling; Artificial intelligence; Class (philosophy); Set (abstract data type); Sampling distribution; Training set; Pattern recognition (psychology); Bandwidth (computing); Mathematics; Statistics","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":[0.009787401,0.001127784,0.001556408,0.002588915,0.001029834,0.001550326,0.002708646,0.001121312,0.00171202],"category_scores_gemma":[0.01469872,0.0005946049,0.001421494,0.001807034,0.001229609,0.002457615,0.002130838,0.002295897,0.0007249273],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009997598,"about_ca_system_score_gemma":0.001391492,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002658276,"about_ca_topic_score_gemma":0.00393508,"domain_scores_codex":[0.995187,0.00191306,0.0002173369,0.0008350042,0.001588925,0.0002587109],"domain_scores_gemma":[0.9946195,0.002425161,0.0004674967,0.001147769,0.001111312,0.0002286799],"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":[0.00062232,0.0006055444,0.0105596,0.0002814705,0.000289252,0.0003080528,0.000756986,0.3499493,0.01431898,0.07021918,0.01047395,0.5416154],"study_design_scores_gemma":[0.0000467178,0.0001424036,0.001153208,0.00002248786,0.00002674996,0.0001487566,0.00008884143,0.9568874,0.003277645,0.03212438,0.006048772,0.00003256781],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.003968132,0.0001072621,0.9946145,0.00008314722,0.0000319449,0.00008451339,0.00006296893,0.000708558,0.0003389093],"genre_scores_gemma":[0.2134397,0.000210247,0.7824909,0.0002988207,0.0002625896,0.0005912922,0.0008429554,0.0003477377,0.001515727],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.009787401,"threshold_uncertainty_score":0.05176133,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1637551525145151,"score_gpt":0.401709392639956,"score_spread":0.2379542401254409,"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."}}