{"id":"W4406518354","doi":"10.1016/j.eswa.2025.126437","title":"Safe deep reinforcement learning for flow control within the Internet of Vehicles","year":2025,"lang":"en","type":"article","venue":"Expert Systems with Applications","topic":"Traffic control and management","field":"Engineering","cited_by":5,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université de Sherbrooke","funders":"Université Mohammed VI Polytechnique","keywords":"Reinforcement learning; Computer science; Artificial intelligence; The Internet; Control (management); Flow (mathematics); Machine learning; World Wide Web; 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":[0.0001393577,0.0001021294,0.0001617278,0.00005039418,0.00007784544,0.00002784342,0.0001707918,0.00002896705,0.000003538004],"category_scores_gemma":[0.000006968736,0.00006800304,0.00003949645,0.0001160065,0.00002748109,0.00002905113,0.00001434674,0.00006388816,0.000005327746],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004256804,"about_ca_system_score_gemma":0.00001322339,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006174827,"about_ca_topic_score_gemma":0.00003726197,"domain_scores_codex":[0.9993832,0.00001447756,0.0002750921,0.0001169989,0.00008859802,0.0001216484],"domain_scores_gemma":[0.9995125,0.0001051371,0.00005894351,0.0002450955,0.000054846,0.0000234978],"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":[0.00002387781,0.00001089385,0.00003042833,0.0001188926,0.000186865,6.795481e-8,0.0005611376,0.9560362,0.000442068,0.03372493,0.002404528,0.006460097],"study_design_scores_gemma":[0.0007054541,0.00002754862,0.00008062206,0.00006058105,0.00002981751,5.233265e-7,0.0006820072,0.8023694,0.0001545918,0.00001917229,0.1958001,0.00007021547],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.000245444,0.001670344,0.9914725,0.0002578789,0.0001269061,0.001869594,0.000001566021,0.0001554601,0.004200338],"genre_scores_gemma":[0.9911125,0.00001446382,0.0008175809,0.00006326114,0.000058648,0.006221466,0.00001082426,0.00001369489,0.001687524],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9908671,"threshold_uncertainty_score":0.2773084,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.005252705768151372,"score_gpt":0.2086016895101681,"score_spread":0.2033489837420167,"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."}}