{"id":"W2947183726","doi":"10.1287/moor.2021.1177","title":"On Linear Programming for Constrained and Unconstrained Average-Cost Markov Decision Processes with Countable Action Spaces and Strictly Unbounded Costs","year":2021,"lang":"en","type":"article","venue":"Mathematics of Operations Research","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Alberta Innovates; Alberta Innovates - Technology Futures; DeepMind; Alberta Machine Intelligence Institute","keywords":"Mathematics; Countable set; Markov decision process; Duality (order theory); Mathematical optimization; State space; Action (physics); Markov kernel; Markov chain; Discrete mathematics; Applied mathematics; Markov process; Markov model; Variable-order Markov model; Statistics","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":[0.0009608432,0.0001293423,0.0002081609,0.0002130498,0.0004932531,0.000856852,0.0002192334,0.00006330857,0.00001189472],"category_scores_gemma":[0.002712241,0.0001053147,0.00001647706,0.0007212295,0.0002403018,0.0004266702,0.000137432,0.0001898997,0.000002039734],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006853649,"about_ca_system_score_gemma":0.0007837786,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001524892,"about_ca_topic_score_gemma":0.0001865762,"domain_scores_codex":[0.9984006,0.0000649227,0.0002980748,0.000315358,0.0006394931,0.0002815055],"domain_scores_gemma":[0.995939,0.002101253,0.00006263101,0.000330847,0.00147538,0.0000909579],"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":[0.0002931662,0.001203603,0.0005941934,0.005102712,0.0003115466,0.00005149548,0.007827589,0.4970565,0.008713468,0.3998643,0.0004401256,0.07854129],"study_design_scores_gemma":[0.001451335,0.0008433194,0.00003690177,0.0006510528,0.00001512668,0.0000704877,0.001822867,0.9825637,0.0106632,0.001251866,0.0004313244,0.0001987952],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.08633459,0.00004553453,0.9115574,0.0003639011,0.00002540769,0.001061943,0.000009453721,0.000032731,0.0005689937],"genre_scores_gemma":[0.4551454,0.00007824663,0.544117,0.00001459798,0.00001311402,0.0001067391,0.00001716528,0.00001236196,0.0004953548],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.4855073,"threshold_uncertainty_score":0.8262644,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07719914338480655,"score_gpt":0.3803864094396801,"score_spread":0.3031872660548735,"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."}}