{"id":"W1554366315","doi":"10.1023/a:1022145020786","title":"Approximate Gradient Methods in Policy-Space Optimization of Markov Reward Processes","year":2003,"lang":"en","type":"article","venue":"Discrete Event Dynamic Systems","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":52,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"","keywords":"Convergence (economics); Variance (accounting); Markov chain; Computer science; Markov process; Gradient descent; Path (computing); Set (abstract data type); Mathematics; Process (computing); Mathematical optimization; Algorithm; Artificial intelligence; Statistics; Artificial neural network","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004817448,0.001255568,0.002799134,0.00116249,0.0007666865,0.001810638,0.001968259,0.002703482,0.002466081],"category_scores_gemma":[0.01970825,0.001543435,0.0007409927,0.001242688,0.002675541,0.002486376,0.002456454,0.002769692,0.0003386377],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00253922,"about_ca_system_score_gemma":0.003166313,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01887139,"about_ca_topic_score_gemma":0.01099634,"domain_scores_codex":[0.9987594,0.0007455999,0.00005247945,0.0001057553,0.000222179,0.0001145852],"domain_scores_gemma":[0.9914162,0.007359692,0.000307921,0.0001682218,0.0005204189,0.0002277148],"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":[0.00005697101,0.00003370088,0.0002008925,0.00006723974,0.00003096208,0.00002044832,0.00004597876,0.9444474,0.000107804,0.04538145,0.0005482123,0.009059083],"study_design_scores_gemma":[0.00000816884,0.000004806346,0.00001595437,0.000003879809,0.000002317862,0.000001275305,0.000002263841,0.9887266,0.00002281328,0.01112045,0.00008958683,0.000002033397],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01095343,0.001028344,0.9852496,0.0006218191,0.00008475014,0.00004371025,0.00003582344,0.0001568213,0.001825712],"genre_scores_gemma":[0.726266,0.001525805,0.2616753,0.0003154683,0.0002383254,0.000540505,0.0002157355,0.0003288473,0.008893954],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01887139,"threshold_uncertainty_score":0.03752309,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01233715985053953,"score_gpt":0.3126734131687594,"score_spread":0.3003362533182199,"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."}}