{"id":"W4386647553","doi":"10.1016/j.compchemeng.2023.108413","title":"Control invariant set enhanced safe reinforcement learning: Improved sampling efficiency, guaranteed stability and robustness","year":2023,"lang":"en","type":"article","venue":"Computers & Chemical Engineering","topic":"Advanced Control Systems Optimization","field":"Engineering","cited_by":9,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada; Alberta Innovates","keywords":"Robustness (evolution); Reinforcement learning; Control theory (sociology); Invariant (physics); Mathematics; Stability (learning theory); Computer science; Mathematical optimization; Engineering; Artificial intelligence; Control (management); Machine learning; Chemistry","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0002589606,0.0003084687,0.0004143977,0.0001244789,0.00006368004,0.00006344932,0.0001874604,0.0001285813,0.00000688017],"category_scores_gemma":[0.0001703683,0.0003412096,0.00006752075,0.0003997621,0.00002969615,0.0001494835,0.00009064817,0.0003287867,0.000005650409],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001776245,"about_ca_system_score_gemma":0.00001301905,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00000450451,"about_ca_topic_score_gemma":1.955145e-7,"domain_scores_codex":[0.9984394,0.00001628948,0.0004712449,0.0003673507,0.0001758296,0.000529843],"domain_scores_gemma":[0.9991994,0.0002941118,0.00005637599,0.0002480144,0.00005351395,0.0001486039],"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":[0.00001135271,0.0000027026,0.000006743031,0.0001256072,0.00002868683,0.000001582962,0.0001312052,0.6370313,0.3622513,0.00003588652,0.000004810883,0.0003688002],"study_design_scores_gemma":[0.001277373,0.00002323439,0.00002927056,0.00009771913,0.00001614773,0.00000566315,0.00002644567,0.9639443,0.03415037,0.000004612861,0.00008634223,0.0003385731],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.202587,0.00007153519,0.7953754,0.00001835579,0.0003411307,0.0003528612,0.000003121328,0.001215568,0.00003505025],"genre_scores_gemma":[0.9921544,0.0000116352,0.007531501,0.00001323027,0.0001183574,0.00006623691,0.00003654528,0.00006312854,0.000005027924],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7895674,"threshold_uncertainty_score":0.999904,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00995526697015245,"score_gpt":0.2032192585718763,"score_spread":0.1932639916017238,"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."}}