{"id":"W3175406278","doi":"10.1109/tai.2021.3074122","title":"Optimal Policy for Bernoulli Bandits: Computation and Algorithm Gauge","year":2021,"lang":"en","type":"article","venue":"IEEE Transactions on Artificial Intelligence","topic":"Advanced Bandit Algorithms Research","field":"Decision Sciences","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Reinforcement learning; Computer science; Bernoulli's principle; Algorithm; Mathematical optimization; Computation; Variety (cybernetics); Approximate Bayesian computation; Time horizon; Thompson sampling; Bayesian probability; Artificial intelligence; Mathematics; Inference","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0009088255,0.0002145378,0.0003087966,0.0005140532,0.0006005295,0.0004215766,0.0003294162,0.0001279614,0.000337555],"category_scores_gemma":[0.0005396174,0.0001986053,0.0001659994,0.001500253,0.0002999241,0.0004150641,0.000006959654,0.0003120035,0.000303222],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001142858,"about_ca_system_score_gemma":0.0002917866,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006321401,"about_ca_topic_score_gemma":0.0001452722,"domain_scores_codex":[0.9967812,0.0001675774,0.0007362015,0.0008005803,0.001046985,0.0004674245],"domain_scores_gemma":[0.9960719,0.002271462,0.0001209988,0.0003893138,0.0009132416,0.0002330658],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00006814166,0.0001710678,6.477033e-7,0.000004534475,0.00001800845,0.000013625,0.0003171574,0.1490886,0.0007488881,0.0007460983,0.0001229779,0.8487002],"study_design_scores_gemma":[0.0001139924,0.0002777654,0.00001916972,0.00002067169,0.00001549999,0.00004564232,0.001478601,0.6159874,0.2882651,0.09066554,0.002834002,0.0002765849],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.003170806,0.00007917414,0.9931659,0.001979131,0.0007448653,0.0003845285,0.0003207098,0.0000651155,0.00008983653],"genre_scores_gemma":[0.8046497,0.0001948853,0.1915577,0.0003382264,0.0005111034,0.0001584349,0.00002009303,0.00005241187,0.002517421],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8484237,"threshold_uncertainty_score":0.809889,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1674809196558598,"score_gpt":0.4598130319407309,"score_spread":0.2923321122848711,"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."}}