{"id":"W2001652071","doi":"10.1016/j.compmedimag.2013.12.003","title":"Ensemble-based hybrid probabilistic sampling for imbalanced data learning in lung nodule CAD","year":2013,"lang":"en","type":"article","venue":"Computerized Medical Imaging and Graphics","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":48,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Alberta","funders":"Alberta Innovates; China Scholarship Council; National Natural Science Foundation of China; Simon Fraser University; University of Alberta","keywords":"Probabilistic logic; Artificial intelligence; Pattern recognition (psychology); Subspace topology; CAD; Computer science; Dimensionality reduction; Sampling (signal processing); Nodule (geology); Computer-aided diagnosis; Probabilistic classification; Statistics; Mathematics; Machine learning; Support vector machine; Computer vision; Naive Bayes classifier; Biology","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":[0.009749478,0.0008364868,0.00252487,0.001586205,0.0009220343,0.001212055,0.002603891,0.001672692,0.001085134],"category_scores_gemma":[0.01467676,0.0007319444,0.001379732,0.001413804,0.0006855943,0.002148026,0.002313119,0.002007207,0.0002863852],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007687745,"about_ca_system_score_gemma":0.0009277831,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004281714,"about_ca_topic_score_gemma":0.004849794,"domain_scores_codex":[0.9966689,0.001696093,0.0002279014,0.0004631133,0.0007356355,0.0002083116],"domain_scores_gemma":[0.9885305,0.007930067,0.0003623146,0.001086041,0.001844228,0.0002468371],"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":[0.0004744643,0.0002632746,0.00680894,0.0001147476,0.0002741088,0.00006894719,0.000168662,0.6244611,0.001974009,0.005707916,0.002878335,0.3568056],"study_design_scores_gemma":[0.000004517643,0.00002093875,0.0002205172,0.000003720889,0.00001008316,0.000009637656,0.000005885547,0.9980029,0.0002347574,0.001369708,0.0001143662,0.000002968923],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03711737,0.0008938058,0.9608816,0.0001927516,0.00006974352,0.00005355307,0.00009447095,0.0003733074,0.0003233574],"genre_scores_gemma":[0.7290345,0.000609027,0.267057,0.0002583986,0.0002591142,0.000212664,0.001000306,0.0001279549,0.001441186],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009749478,"threshold_uncertainty_score":0.05156076,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03378149451165669,"score_gpt":0.3051422654233574,"score_spread":0.2713607709117007,"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."}}