{"id":"W3103712156","doi":"10.48550/arxiv.2103.12243","title":"Adaptive Importance Sampling for Finite-Sum Optimization and Sampling with Decreasing Step-Sizes","year":2021,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Regret; Variance reduction; Estimator; Sampling (signal processing); Mathematical optimization; Convergence (economics); Variance (accounting); Computer science; Importance sampling; Rate of convergence; Adaptive sampling; Mathematics; Algorithm; Key (lock); Statistics; Machine learning; Monte Carlo method","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":[],"consensus_categories":[],"category_scores_codex":[0.0001320722,0.0001583813,0.000170736,0.000126757,0.0002566628,0.0001190707,0.0002695569,0.00006482469,0.000006868799],"category_scores_gemma":[0.0001624437,0.0001755689,0.00004242514,0.0007029948,0.00006844606,0.000605124,0.0001521072,0.00008078869,6.526907e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007837049,"about_ca_system_score_gemma":0.0001118407,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001082665,"about_ca_topic_score_gemma":0.00001709215,"domain_scores_codex":[0.9988859,0.00003464277,0.0001393209,0.0006293858,0.00006591179,0.0002448422],"domain_scores_gemma":[0.9986042,0.0004545908,0.0001457234,0.0003555183,0.0003311398,0.0001088764],"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":[0.00003665617,0.00002752721,0.002042555,0.00001276592,0.00002665956,0.00002115807,0.0001172047,0.7876565,0.00002642471,0.2095455,0.000006503066,0.0004805682],"study_design_scores_gemma":[0.0005747449,0.0001024772,0.0002382932,0.00007294647,0.00003331715,0.00001436697,0.0002121786,0.9906244,0.0003227345,0.007526002,0.00003900887,0.0002395682],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01119303,0.00006478715,0.9879469,0.00004932187,0.00005874456,0.0002464865,0.00000624927,0.0002424874,0.0001920009],"genre_scores_gemma":[0.443673,0.00003283222,0.5561512,0.00005647326,0.00001236874,0.000001645942,0.00000838264,0.00001037225,0.00005376183],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.43248,"threshold_uncertainty_score":0.7159491,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09071579776244318,"score_gpt":0.2069464663268825,"score_spread":0.1162306685644394,"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."}}