{"id":"W4318604573","doi":"10.1109/ssci51031.2022.10022190","title":"Training and Testing Cascades for Imbalanced Data Classification","year":2022,"lang":"en","type":"article","venue":"2022 IEEE Symposium Series on Computational Intelligence (SSCI)","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Regina","funders":"","keywords":"Cascade; Computer science; Artificial intelligence; Training set; Machine learning; Data modeling; Data mining; Data set; Class (philosophy); Set (abstract data type); Synthetic data; Test data; Recall rate; Pattern recognition (psychology); Database; Engineering","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":[0.007621484,0.002088299,0.001501076,0.002078569,0.001164995,0.0009869138,0.00289645,0.001501235,0.002922348],"category_scores_gemma":[0.02183869,0.0007882854,0.001013775,0.0009726672,0.000769843,0.004194192,0.002159705,0.002486309,0.00182249],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009552904,"about_ca_system_score_gemma":0.001385316,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002272583,"about_ca_topic_score_gemma":0.00429502,"domain_scores_codex":[0.9967321,0.00087755,0.0002465284,0.0008634963,0.001033044,0.0002473354],"domain_scores_gemma":[0.9876744,0.005624623,0.0008945186,0.003177739,0.002087257,0.0005414682],"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":[0.002133553,0.002551196,0.04993428,0.0003467902,0.0003714502,0.0007510295,0.0007560701,0.1410042,0.02770375,0.008570882,0.01566347,0.7502133],"study_design_scores_gemma":[0.00005850881,0.0003679645,0.002580175,0.00002946276,0.00004994997,0.0001793682,0.0001055062,0.9747217,0.01139739,0.008375159,0.002114163,0.00002071248],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2205865,0.0008707326,0.7612559,0.0007490281,0.0004362763,0.001816171,0.0007026222,0.008166374,0.005416333],"genre_scores_gemma":[0.6717175,0.0002535847,0.3221374,0.0003463469,0.0002217079,0.0009406859,0.001309194,0.0002738873,0.00279966],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007621484,"threshold_uncertainty_score":0.04030675,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1728539108943217,"score_gpt":0.3439910303025261,"score_spread":0.1711371194082045,"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."}}