{"id":"W2904232271","doi":"10.1609/aaai.v33i01.33016112","title":"Leveraging Observations in Bandits: Between Risks and Benefits","year":2019,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference 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":"","keywords":"Regret; Multi-armed bandit; Leverage (statistics); Computer science; Optimism; Context (archaeology); Thompson sampling; Order (exchange); Task (project management); Artificial intelligence; Machine learning; Dependency (UML); Economics; 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":[],"consensus_categories":[],"category_scores_codex":[0.002039528,0.0001862796,0.0003736943,0.0003770996,0.0001571374,0.0003035532,0.001442338,0.00009726139,0.0002319577],"category_scores_gemma":[0.003439764,0.0001290537,0.00007187748,0.001482867,0.0002646507,0.0006039467,0.000451759,0.0004563788,0.0002127373],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005554901,"about_ca_system_score_gemma":0.00008415789,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009205228,"about_ca_topic_score_gemma":0.00003671005,"domain_scores_codex":[0.9967798,0.00003467439,0.0008355797,0.0006261779,0.001337555,0.0003862093],"domain_scores_gemma":[0.997282,0.0009888215,0.0003749671,0.0003171545,0.000936763,0.0001002666],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001346376,0.000140309,0.4626327,0.00003936654,0.00002102021,6.964164e-7,0.003846914,0.002854859,0.008324088,0.1299295,0.0001384219,0.3919376],"study_design_scores_gemma":[0.0001035382,0.0002002017,0.2691953,0.0003522986,0.00000885749,0.000002189592,0.005174913,0.03458942,0.126291,0.5634787,0.0002654003,0.0003381107],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9934001,0.00004032602,0.001348623,0.002202028,0.0001857869,0.0004979267,0.00001542116,0.00002021882,0.002289593],"genre_scores_gemma":[0.9986663,0.00004481131,0.0006236008,0.00006459908,0.00005792549,0.00001675731,6.365495e-7,0.00001217643,0.0005132044],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4335493,"threshold_uncertainty_score":0.5262656,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.534976651882224,"score_gpt":0.4463792313613104,"score_spread":0.08859742052091357,"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."}}