{"id":"W3208237615","doi":"10.48550/arxiv.2110.15412","title":"Stochastic Mirror Descent: Convergence Analysis and Adaptive Variants via the Mirror Stochastic Polyak Stepsize","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Stochastic gradient descent; Bounded function; Convergence (economics); Adaptive stepsize; Regular polygon; Mathematical optimization; Descent (aeronautics); Mathematics; Convex function; Stochastic optimization; Computer science; Gradient descent; Applied mathematics; Artificial intelligence; Numerical analysis; Mathematical analysis; Artificial neural network; Physics; Geometry","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"],"consensus_categories":[],"category_scores_codex":[0.0004132972,0.0006306052,0.0007558206,0.0007353335,0.0004286541,0.0003316189,0.002456968,0.0003573915,0.00009516333],"category_scores_gemma":[0.0001749415,0.0006204565,0.0003991816,0.002928538,0.0004289525,0.000486289,0.003070861,0.000757922,0.00002116957],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002839212,"about_ca_system_score_gemma":0.0003353845,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001191911,"about_ca_topic_score_gemma":0.0003039615,"domain_scores_codex":[0.9962558,0.0003680867,0.0004440717,0.00200407,0.0003023667,0.0006256484],"domain_scores_gemma":[0.995961,0.0005212044,0.0005920087,0.002048365,0.0005323946,0.000345058],"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":[0.0002542379,0.0006888229,0.002044737,0.0001732374,0.005440278,0.0007685298,0.005137863,0.7803977,0.0003069553,0.2036338,0.00008871978,0.001065216],"study_design_scores_gemma":[0.000405892,0.00008622175,0.006139267,0.0001238712,0.001255586,0.00002057974,0.0002728391,0.9819127,0.00004833411,0.009072826,0.000001409979,0.000660525],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03236873,0.0001697729,0.9653866,0.0001369096,0.0006126386,0.000838195,0.00004076261,0.0003743152,0.0000720754],"genre_scores_gemma":[0.9682632,0.00004330883,0.03113492,0.0001377843,0.00001614642,0.0000120403,0.00002687619,0.00003231902,0.0003333908],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9358945,"threshold_uncertainty_score":0.9996247,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05651900974294029,"score_gpt":0.1963108193767179,"score_spread":0.1397918096337776,"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."}}