{"id":"W4281561301","doi":"10.48550/arxiv.2203.09612","title":"Risk-Averse Markov Decision Processes through a Distributional Lens","year":2022,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Risk and Portfolio Optimization","field":"Decision Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Markov decision process; Dynamic programming; Mathematical optimization; Limit (mathematics); Regular polygon; Markov process; Invariant (physics); Markov chain; Risk measure; Computer science; Econometrics; Order (exchange); Mathematics; Economics; Statistics; Machine learning","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001538667,0.0004230328,0.0005394028,0.0004291315,0.0008233829,0.0002858381,0.002048939,0.0003540622,0.004428119],"category_scores_gemma":[0.003045155,0.0004116915,0.0004066236,0.002554563,0.0002227593,0.000863441,0.002396525,0.0009409865,0.0004363482],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000380848,"about_ca_system_score_gemma":0.0008681258,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003224286,"about_ca_topic_score_gemma":0.0001500957,"domain_scores_codex":[0.9957636,0.0004849666,0.000643961,0.00180467,0.000869624,0.0004331529],"domain_scores_gemma":[0.9946888,0.001596355,0.001115707,0.001505258,0.0009281711,0.0001657044],"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":[0.0003968182,0.0002148186,0.09621166,0.00001665727,0.00008443431,0.0002647888,0.0003274888,0.8616864,8.193339e-7,0.01424817,0.024282,0.002265915],"study_design_scores_gemma":[0.001486366,0.0001552898,0.02353868,0.00008425426,0.0003663999,0.00001936583,0.001757218,0.1270916,0.00003325245,0.5366469,0.3074993,0.001321384],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4798423,0.0002179939,0.5005875,0.0001134249,0.001502991,0.0005315332,0.001732665,0.0001816247,0.01529005],"genre_scores_gemma":[0.9841583,0.007181786,0.00198533,0.00008485831,0.0001401571,0.000003778393,0.000455055,0.00002864588,0.005962028],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7345949,"threshold_uncertainty_score":0.9998335,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.148469092824942,"score_gpt":0.2724317286602049,"score_spread":0.1239626358352629,"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."}}