{"id":"W2062227835","doi":"10.1109/icassp.2013.6639346","title":"Improving deep neural networks for LVCSR using rectified linear units and dropout","year":2013,"lang":"en","type":"article","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":1281,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Dropout (neural networks); Artificial neural network; Speech recognition; Artificial intelligence; Sigmoid function; Discriminative model; Deep neural networks; Task (project management); Vocabulary; Deep learning; Mixture model; Hidden Markov model; Time delay neural network; Bayesian probability; Pattern recognition (psychology); 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":[0.002555288,0.002196582,0.001022506,0.0008457405,0.0006046871,0.0009731093,0.001693571,0.001547885,0.006583173],"category_scores_gemma":[0.005527613,0.0007281979,0.0009515456,0.0007407653,0.0005077788,0.002338533,0.001190327,0.00333959,0.003333926],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001666879,"about_ca_system_score_gemma":0.001470251,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02303117,"about_ca_topic_score_gemma":0.03374742,"domain_scores_codex":[0.999082,0.0001901563,0.00007255267,0.0002367718,0.0003019021,0.0001166822],"domain_scores_gemma":[0.9988064,0.0005776086,0.00006854966,0.0001678975,0.0003395939,0.00003999847],"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":[0.0003462687,0.000230877,0.001197521,0.0002064745,0.0001566436,0.0001689383,0.0001289763,0.3860774,0.02378022,0.004353059,0.009505271,0.5738485],"study_design_scores_gemma":[0.00001970191,0.00007813721,0.0003080294,0.00001791244,0.00002275298,0.00003553672,0.00001440596,0.9854512,0.01079618,0.001231429,0.002008651,0.0000160608],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03877688,0.002205609,0.9407361,0.0004662909,0.0002238933,0.0001262378,0.0004304609,0.01308756,0.003946934],"genre_scores_gemma":[0.4202078,0.001042487,0.5617856,0.000572659,0.0001302954,0.0002781184,0.002304496,0.0009571741,0.01272141],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02303117,"threshold_uncertainty_score":0.04579425,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0442634602867148,"score_gpt":0.2510945759761163,"score_spread":0.2068311156894015,"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."}}