{"id":"W4384575244","doi":"10.23952/jano.5.2023.2.04","title":"IdentifyMix: An efficient two-stage learning approach to combating label noise","year":2023,"lang":"en","type":"article","venue":"Journal of Applied and Numerical Optimization","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Science Foundation of Fujian Province; National Natural Science Foundation of China","keywords":"Artificial intelligence; Computer science; Machine learning; Margin (machine learning); Noise (video); Generalization; Sample (material); Artificial neural network; Selection (genetic algorithm); Supervised learning; Process (computing); Pattern recognition (psychology); Deep learning; Stability (learning theory); Mathematics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008094303,0.000100322,0.0001762861,0.0002244029,0.0002043503,0.0002500853,0.0003007627,0.00003715295,0.000003568812],"category_scores_gemma":[0.0001007672,0.0000856493,0.00002526537,0.0007597396,0.00001524922,0.0002441253,0.0001219756,0.0002514662,0.00001283428],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002365501,"about_ca_system_score_gemma":0.00002559703,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00000475079,"about_ca_topic_score_gemma":5.127884e-8,"domain_scores_codex":[0.9988527,0.00007831474,0.0003405011,0.0002324391,0.0003263784,0.0001697123],"domain_scores_gemma":[0.9991852,0.00007171869,0.0003049032,0.0001617231,0.00009664761,0.0001798176],"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":[0.00002190058,0.000103976,0.0001038074,0.000009678008,0.000005821869,0.000001559488,0.0008110933,0.9697666,0.0009849477,0.009652938,0.00004729423,0.01849035],"study_design_scores_gemma":[0.0004462821,0.0001389369,0.0009949395,0.000009257437,0.000007025433,0.000008300688,0.0001695396,0.9974414,0.00007342035,0.00005982324,0.0005482484,0.0001028396],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03282287,0.00001010195,0.964647,0.0002896892,0.00008738841,0.00008777161,5.626941e-7,0.00009860006,0.001955996],"genre_scores_gemma":[0.7188635,0.00001221936,0.2808405,0.0001121732,0.00007304284,0.000003853795,0.0000203603,0.0000100507,0.00006431306],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6860406,"threshold_uncertainty_score":0.3492677,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02130369396280206,"score_gpt":0.2830791325096442,"score_spread":0.2617754385468421,"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."}}