{"id":"W2920229473","doi":"10.48550/arxiv.1903.01026","title":"Learning Modular Safe Policies in the Bandit Setting with Application to Adaptive Clinical Trials","year":2019,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Advanced Bandit Algorithms Research","field":"Decision Sciences","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Regret; Robustness (evolution); Outlier; Computer science; Modular design; Mathematical optimization; Measure (data warehouse); Estimator; Flexibility (engineering); Suite; Artificial intelligence; Machine learning; Mathematics; Data mining; 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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.02344378,0.0003365663,0.001011833,0.0007920297,0.0002545764,0.0002672101,0.002670631,0.0003561603,0.00006215647],"category_scores_gemma":[0.008800589,0.0002278383,0.0003239884,0.001983374,0.0002567128,0.0003054026,0.001418051,0.001965854,0.0006093191],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002613019,"about_ca_system_score_gemma":0.0003751626,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003066083,"about_ca_topic_score_gemma":0.0002181022,"domain_scores_codex":[0.9909417,0.004909497,0.001042558,0.001700557,0.0008666996,0.0005389628],"domain_scores_gemma":[0.9849326,0.01160826,0.0009994823,0.001685604,0.0005798647,0.0001941617],"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":[0.000411392,0.00005956687,0.01870765,0.00000679732,0.00004642137,0.00005916879,0.0008623391,0.9696072,0.000008476795,0.001484804,0.000237103,0.00850911],"study_design_scores_gemma":[0.001731124,0.0006337645,0.05629474,0.0001558426,0.00008465868,0.000006121857,0.01214998,0.8786664,0.00004058297,0.04222046,0.007326329,0.0006899819],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3436402,0.000025756,0.652779,0.0004949184,0.0001501907,0.00158899,0.00002674009,0.00003876885,0.001255438],"genre_scores_gemma":[0.9960042,0.0000642271,0.0009784108,0.0001545299,0.000255202,0.00001091233,0.0000164887,0.00002576816,0.00249027],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.652364,"threshold_uncertainty_score":0.9995487,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4112833942530587,"score_gpt":0.4118405645397245,"score_spread":0.0005571702866657979,"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."}}