{"id":"W2970168598","doi":"","title":"Surrogate Objectives for Batch Policy Optimization in One-step Decision Making","year":2019,"lang":"en","type":"article","venue":"Neural Information Processing Systems","topic":"Advanced Bandit Algorithms Research","field":"Decision Sciences","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Mathematical optimization; Entropy (arrow of time); Artificial intelligence; Multi-armed bandit; Constrained optimization problem; Action selection; Action (physics); Optimization problem; State (computer science); Surrogate model; Principle of maximum entropy; Machine learning; Mathematics; Algorithm; Regret; Psychology","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":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.002499276,0.0001911412,0.0003873786,0.001421658,0.0002350704,0.001441036,0.0005957565,0.0001410922,0.00003305998],"category_scores_gemma":[0.003984449,0.0001551558,0.00007503654,0.002149529,0.00004658812,0.007755655,0.0001175971,0.0002048562,0.0001912329],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003051857,"about_ca_system_score_gemma":0.0002705393,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005038744,"about_ca_topic_score_gemma":0.00001355224,"domain_scores_codex":[0.995966,0.0001408866,0.001373479,0.0003486334,0.001708416,0.0004625818],"domain_scores_gemma":[0.9963411,0.001208364,0.0007218799,0.0003758853,0.001274584,0.0000781864],"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":[0.0001270401,0.00001570159,0.001104761,0.0001121106,0.00000241504,2.922987e-7,0.0009601046,0.8006089,0.00002597876,0.0001402936,0.00009718435,0.1968052],"study_design_scores_gemma":[0.000939889,0.00008547748,0.001431157,0.0002959299,0.000001661132,0.00001060678,0.00225314,0.9918358,0.00007070589,0.000826301,0.00206976,0.0001796144],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04015015,0.0001343173,0.9554669,0.0002528422,0.0006922986,0.001608467,0.00002528587,0.00009816168,0.001571635],"genre_scores_gemma":[0.9881794,0.000006473802,0.01097521,0.0001109995,0.0001444943,0.0001218422,0.00002375631,0.00001780909,0.0004199812],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9480293,"threshold_uncertainty_score":0.9995956,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07992803755470389,"score_gpt":0.4233275306939406,"score_spread":0.3433994931392367,"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."}}