{"id":"W3211602915","doi":"10.1007/s11227-021-04177-6","title":"A binary PSO-based ensemble under-sampling model for rebalancing imbalanced training data","year":2021,"lang":"en","type":"article","venue":"The Journal of Supercomputing","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":32,"is_retracted":false,"has_abstract":false,"ca_institutions":"Lakehead University","funders":"","keywords":"Computer science; Artificial intelligence; Ensemble learning; Machine learning; Sampling (signal processing); Selection (genetic algorithm); Binary number; Binary classification; Data mining; Pattern recognition (psychology); Support vector machine; Mathematics","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":[0.002172574,0.0008188144,0.00233869,0.000803353,0.0005722275,0.001079023,0.002450587,0.001660655,0.002232677],"category_scores_gemma":[0.006074038,0.0005842202,0.0008627439,0.001080135,0.000627206,0.001680174,0.001409019,0.001618336,0.0005988973],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005679625,"about_ca_system_score_gemma":0.001165121,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007631116,"about_ca_topic_score_gemma":0.007560675,"domain_scores_codex":[0.9990757,0.0002771724,0.00005249807,0.0002228088,0.0002527922,0.0001189908],"domain_scores_gemma":[0.9981179,0.0009213274,0.0001342224,0.0001809248,0.000526167,0.0001194946],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002667598,0.000169417,0.001434722,0.00007233072,0.0001016186,0.00005434403,0.00006651432,0.8084112,0.001882513,0.00557362,0.003833462,0.1781335],"study_design_scores_gemma":[0.000004067228,0.00001416446,0.00004489619,0.000002155298,0.000004354425,0.000004904611,0.000001662367,0.9993845,0.00008293242,0.0003566877,0.00009808577,0.000001514605],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02205984,0.0008051249,0.9740661,0.0003078925,0.0002299289,0.0000664929,0.00008191045,0.0005499176,0.001832754],"genre_scores_gemma":[0.7152858,0.0005560167,0.275139,0.0006394766,0.000406899,0.0002861694,0.0005643183,0.0002050545,0.00691722],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007631116,"threshold_uncertainty_score":0.01517344,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2111408925881917,"score_gpt":0.3510560927424298,"score_spread":0.1399152001542381,"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."}}