{"id":"W104184427","doi":"","title":"On the importance of initialization and momentum in deep learning","year":2013,"lang":"en","type":"article","venue":"","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":3536,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Initialization; Momentum (technical analysis); Computer science; Recurrent neural network; Gradient descent; Stochastic gradient descent; Deep learning; Artificial intelligence; Deep neural networks; Artificial neural network; Schedule; Machine learning; Mathematical optimization; Algorithm; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007742895,0.001373461,0.001018042,0.0007394888,0.001111797,0.002136893,0.0009646331,0.002123482,0.001328377],"category_scores_gemma":[0.05319958,0.001073979,0.0004071493,0.001009419,0.003864856,0.006069802,0.002529216,0.005307974,0.0006796541],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001177029,"about_ca_system_score_gemma":0.001271661,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00442008,"about_ca_topic_score_gemma":0.003668925,"domain_scores_codex":[0.997547,0.001410128,0.0001420973,0.000345857,0.0004301381,0.0001248564],"domain_scores_gemma":[0.975794,0.01992361,0.0008815986,0.001383839,0.001604813,0.0004121189],"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":[0.000707749,0.0001207109,0.005030699,0.0004025294,0.00008804473,0.0003536114,0.0003935376,0.50895,0.006347536,0.2259366,0.00722402,0.2444449],"study_design_scores_gemma":[0.00003981365,0.0001248487,0.0009039799,0.0001639114,0.00002844035,0.00008689453,0.00003119025,0.8940015,0.003321924,0.09868295,0.002565888,0.00004860476],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03739695,0.01226426,0.9313169,0.003665625,0.0007005526,0.00007960347,0.00006379918,0.001099236,0.01341302],"genre_scores_gemma":[0.7206132,0.01025416,0.2616054,0.0008957024,0.001170595,0.0001728285,0.0001879182,0.0007382997,0.004361901],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007742895,"threshold_uncertainty_score":0.04094887,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01169313606498904,"score_gpt":0.2239115137179905,"score_spread":0.2122183776530014,"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."}}