{"id":"W2955301088","doi":"10.1109/ccgrid.2019.00092","title":"A Performance Improvement Approach for Second-Order Optimization in Large Mini-batch Training","year":2019,"lang":"en","type":"article","venue":"","topic":"Blind Source Separation Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Core Research for Evolutional Science and Technology; Canadian Institute for Advanced Research; Nvidia","keywords":"Overfitting; Computer science; Generalization; Regularization (linguistics); Artificial intelligence; Smoothing; Artificial neural network; Deep learning; Training (meteorology); Deep neural networks; Training set; Machine learning; Algorithm; Mathematical optimization; Mathematics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003176333,0.002077454,0.001436283,0.0008818822,0.0006452971,0.0009472629,0.002119655,0.001681624,0.003934054],"category_scores_gemma":[0.006665581,0.0007985975,0.00158492,0.0007426855,0.00143822,0.001662362,0.001764957,0.00358705,0.001764304],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001264487,"about_ca_system_score_gemma":0.002465554,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004639635,"about_ca_topic_score_gemma":0.005127834,"domain_scores_codex":[0.9987551,0.000354576,0.00008382366,0.0002235336,0.0004724898,0.0001105926],"domain_scores_gemma":[0.9978356,0.001028243,0.0001314827,0.0004189237,0.0004827456,0.0001030681],"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":[0.0005897248,0.0003078877,0.001047604,0.0004156107,0.0001934455,0.0001648509,0.0003062821,0.631973,0.03394249,0.05254804,0.009881304,0.2686297],"study_design_scores_gemma":[0.000009909491,0.0000458752,0.00009012869,0.000005264578,0.000007047122,0.00002488627,0.000003733747,0.9939218,0.002613728,0.002350885,0.0009183834,0.000008463809],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.002461542,0.0001788311,0.9956654,0.0001136641,0.00005533367,0.00003847171,0.0000200212,0.0006860425,0.0007807609],"genre_scores_gemma":[0.1417056,0.0004455204,0.8491372,0.0003150674,0.0001939728,0.0004103055,0.0002685669,0.0006645243,0.006859312],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004639635,"threshold_uncertainty_score":0.01679826,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01947409969596551,"score_gpt":0.2522402888070221,"score_spread":0.2327661891110566,"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."}}