{"id":"W1566313113","doi":"10.3233/ida-140686","title":"Hybrid probabilistic sampling with random subspace for imbalanced data learning","year":2014,"lang":"en","type":"article","venue":"Intelligent Data Analysis","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Alberta Innovates; China Scholarship Council","keywords":"Subspace topology; Probabilistic logic; Sampling (signal processing); Artificial intelligence; Computer science; Pattern recognition (psychology); Machine learning; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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":[0.007610859,0.001092668,0.00269214,0.00217358,0.001034992,0.001477916,0.003161105,0.001436398,0.001827924],"category_scores_gemma":[0.01315385,0.0006583729,0.00163065,0.002906554,0.001718988,0.003149765,0.004106917,0.001995746,0.001115541],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008273638,"about_ca_system_score_gemma":0.001693087,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001744139,"about_ca_topic_score_gemma":0.002054589,"domain_scores_codex":[0.9915127,0.003884041,0.0003611181,0.001114619,0.002785881,0.0003416584],"domain_scores_gemma":[0.9938413,0.002707133,0.0004360957,0.001597679,0.001132024,0.0002857583],"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":[0.0005230897,0.0003499985,0.003745934,0.0002223801,0.0003188415,0.0001274402,0.0002532434,0.4033127,0.007408777,0.0700656,0.005532429,0.5081396],"study_design_scores_gemma":[0.00001718058,0.00004574079,0.0001917538,0.000004363761,0.000008867671,0.00003553217,0.000009897329,0.9846761,0.001036792,0.01280288,0.001157847,0.00001313175],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002369235,0.0001125326,0.9968791,0.00005974748,0.00002260702,0.00003095467,0.00002180574,0.0003087372,0.0001953705],"genre_scores_gemma":[0.2003603,0.0003148505,0.7960692,0.0002574209,0.0002722334,0.0003631901,0.0005542804,0.0001873175,0.001621219],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007610859,"threshold_uncertainty_score":0.04025054,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08624709222993643,"score_gpt":0.333776344110833,"score_spread":0.2475292518808965,"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."}}