{"id":"W4206630568","doi":"10.1109/bigdata52589.2021.9671816","title":"MixNN: Combating Noisy Labels in Deep Learning by Mixing with Nearest Neighbors","year":2021,"lang":"en","type":"article","venue":"2021 IEEE International Conference on Big Data (Big Data)","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University","funders":"","keywords":"Overfitting; Computer science; Artificial intelligence; Pattern recognition (psychology); Machine learning; Noise (video); Deep learning; Transfer of learning; Ground truth; Representation (politics); Artificial neural network; Image (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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0008833889,0.000295578,0.0002884918,0.0002074319,0.0002157373,0.001180207,0.005283108,0.0001046436,0.0001531337],"category_scores_gemma":[0.00103322,0.0002867221,0.00002301786,0.0006513114,0.00007992564,0.001722333,0.00213881,0.0008701812,0.000183202],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008976681,"about_ca_system_score_gemma":0.0003825906,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005754101,"about_ca_topic_score_gemma":0.001995293,"domain_scores_codex":[0.9961815,0.000368579,0.0005005599,0.001608419,0.0009354804,0.0004054143],"domain_scores_gemma":[0.9958422,0.0003715259,0.0003359464,0.002984504,0.0003211861,0.0001446508],"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":[0.00006393323,0.0004060942,0.01992664,0.00004048997,0.0001172363,0.0002908801,0.0003734295,0.0008018875,0.008576834,0.01828436,0.008602978,0.9425153],"study_design_scores_gemma":[0.001243661,0.0001123506,0.006865844,0.0004542428,0.00001705663,0.00006187885,0.0003576124,0.8698495,0.001068934,0.0002205159,0.1191166,0.0006318489],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02869965,0.0003318385,0.9289626,0.01569127,0.004379868,0.0003421392,0.003848758,0.0002760796,0.01746782],"genre_scores_gemma":[0.9444745,0.0003530864,0.01256926,0.0005836203,0.0005453682,0.00001749688,0.0404548,0.00003062953,0.0009712294],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9418834,"threshold_uncertainty_score":0.9999585,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1969339514472645,"score_gpt":0.3369666703159024,"score_spread":0.1400327188686379,"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."}}