{"id":"W3118438270","doi":"10.1214/23-aos2315","title":"Relaxing the i.i.d. assumption: Adaptively minimax optimal regret via root-entropic regularization","year":2023,"lang":"en","type":"article","venue":"The Annals of Statistics","topic":"Advanced Bandit Algorithms Research","field":"Decision Sciences","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; University of Toronto","funders":"","keywords":"Regret; Minimax; Mathematics; Mathematical optimization; Regularization (linguistics); Entropy (arrow of time); Adversarial system; Computer science; Artificial intelligence; Statistics","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":[0.005083699,0.00145625,0.001739217,0.0005850055,0.0006319035,0.00153704,0.002191953,0.001863765,0.001175664],"category_scores_gemma":[0.02088718,0.0006401641,0.0008819616,0.0005980608,0.00310702,0.002643517,0.002852336,0.004152127,0.0002923083],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001529254,"about_ca_system_score_gemma":0.001312475,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001746072,"about_ca_topic_score_gemma":0.001372122,"domain_scores_codex":[0.9976035,0.001108708,0.0000786049,0.0005248206,0.0004508356,0.0002336198],"domain_scores_gemma":[0.9887885,0.008262492,0.001209908,0.0009437317,0.0004364849,0.000358743],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000123453,0.00007014453,0.0008283609,0.00006596385,0.00006315733,0.0001524748,0.00009024104,0.8866391,0.001850717,0.09995747,0.001168433,0.008990466],"study_design_scores_gemma":[0.000008951442,0.00003158278,0.0001011486,0.000007731254,0.000005198797,0.00001754315,0.00000722187,0.9677125,0.0004052306,0.03154456,0.0001503273,0.000007945602],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03576383,0.0002947712,0.959896,0.0009047092,0.0000391391,0.0000431172,0.00007518783,0.0001739117,0.002809354],"genre_scores_gemma":[0.9126633,0.0003898609,0.08210339,0.0006139895,0.0001482181,0.0001591171,0.0001290806,0.000108888,0.003684248],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005083699,"threshold_uncertainty_score":0.02688545,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3214253860286645,"score_gpt":0.4722292110110216,"score_spread":0.1508038249823571,"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."}}