{"id":"W2963923407","doi":"","title":"Addressing Function Approximation Error in Actor-Critic Methods","year":2018,"lang":"en","type":"article","venue":"UvA-DARE (University of Amsterdam)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":414,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Reinforcement learning; Computer science; Bellman equation; Suite; Function (biology); Function approximation; Value (mathematics); Limit (mathematics); Artificial intelligence; Approximation algorithm; Temporal difference learning; Mathematical optimization; Machine learning; Algorithm; Artificial neural network; Mathematics","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":[0.0004695347,0.0001167391,0.0001883031,0.0003085093,0.0001965967,0.000061242,0.0006376738,0.00009508116,0.0001140939],"category_scores_gemma":[0.00006597937,0.0001472616,0.00006337802,0.0005404582,0.0001573545,0.001170831,0.0003302252,0.0001625214,0.0000588998],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001143272,"about_ca_system_score_gemma":0.00005961961,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009073262,"about_ca_topic_score_gemma":0.00003730231,"domain_scores_codex":[0.998826,0.0002086075,0.0001579347,0.0003034289,0.0002640247,0.0002400642],"domain_scores_gemma":[0.9990621,0.00009876372,0.0001766799,0.0004361766,0.0001641826,0.00006210866],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004657088,0.0004408836,0.01589933,0.001021615,0.0002674565,0.000107197,0.06703836,0.06675688,0.04366093,0.05056656,0.004253956,0.7495211],"study_design_scores_gemma":[0.001074633,0.0004323008,0.04460939,0.0002488052,0.00003262333,0.000007439533,0.002492132,0.9436291,0.001198327,0.0009356977,0.004993854,0.0003456795],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03467278,0.00001569866,0.9586182,0.0003059115,0.0003963739,0.0001378391,7.969048e-7,0.0001037584,0.005748617],"genre_scores_gemma":[0.7351088,0.000002276131,0.2641078,0.00007124237,0.00003908871,1.837117e-7,0.000005780023,0.000006758804,0.0006580958],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8768722,"threshold_uncertainty_score":0.6005155,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07700182874909071,"score_gpt":0.3215285431639732,"score_spread":0.2445267144148824,"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."}}