{"id":"W2100220834","doi":"10.1109/icassp.2005.1415060","title":"Discriminative Training of CDHMMs for Maximum Relative Separation Margin","year":2006,"lang":"en","type":"article","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":24,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University","funders":"","keywords":"Discriminative model; Minimax; Computer science; Margin (machine learning); Hidden Markov model; Probabilistic logic; Pattern recognition (psychology); Artificial intelligence; Separation (statistics); Reduction (mathematics); Optimization problem; Mathematical optimization; Word error rate; Mathematics; Algorithm; Machine learning","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":[0.001633031,0.0006890495,0.0008554445,0.0003911434,0.0003533205,0.0004393043,0.001310263,0.0009841988,0.00185008],"category_scores_gemma":[0.006118317,0.0005350907,0.0003223962,0.0003896402,0.0007614212,0.001097615,0.001297533,0.001514092,0.000820948],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004626349,"about_ca_system_score_gemma":0.0007492726,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0010637,"about_ca_topic_score_gemma":0.001688229,"domain_scores_codex":[0.99898,0.0004144871,0.00004542149,0.0002340578,0.0002550559,0.0000709811],"domain_scores_gemma":[0.9980685,0.001241669,0.0001537347,0.0002391104,0.0002349536,0.00006196438],"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":[0.000330835,0.0001445667,0.001288227,0.0001829834,0.00007010851,0.00007072876,0.0001357559,0.3539611,0.03710776,0.01604481,0.002578993,0.5880842],"study_design_scores_gemma":[0.00001041348,0.00003715514,0.0002591415,0.000006067087,0.000005084386,0.00003568979,0.000006765885,0.9892403,0.006820847,0.0028643,0.0007064823,0.000007592739],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.006622712,0.00009333928,0.9926049,0.00005359385,0.00001213878,0.00001260326,0.00001216837,0.0002734873,0.0003149621],"genre_scores_gemma":[0.4437044,0.0001360856,0.5526703,0.0002642963,0.00006993298,0.000169852,0.0002593288,0.0002135872,0.002512345],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00185008,"threshold_uncertainty_score":0.008636415,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05021710298073339,"score_gpt":0.2949986045797244,"score_spread":0.244781501598991,"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."}}