{"id":"W2117499988","doi":"","title":"The Difficulty of Training Deep Architectures and the Effect of Unsupervised Pre-Training","year":2009,"lang":"en","type":"article","venue":"International Conference on Artificial Intelligence and Statistics","topic":"Generative Adversarial Networks and Image Synthesis","field":"Computer Science","cited_by":323,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Artificial intelligence; Computer science; Initialization; Robustness (evolution); Deep learning; Training (meteorology); Machine learning; Deep neural networks; Unsupervised learning; Training set","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":[0.004570615,0.001691648,0.001056452,0.000330973,0.0007126205,0.0009829716,0.001545149,0.002213132,0.002451544],"category_scores_gemma":[0.04140577,0.001007346,0.0006096715,0.0005230665,0.0022897,0.003816698,0.002002061,0.004173051,0.0004233043],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007688448,"about_ca_system_score_gemma":0.0008239233,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002482794,"about_ca_topic_score_gemma":0.005429824,"domain_scores_codex":[0.9980217,0.00108524,0.0001123616,0.0003118325,0.0002947183,0.0001742695],"domain_scores_gemma":[0.974034,0.02031298,0.001376273,0.003289734,0.000660355,0.000326675],"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":[0.0007030441,0.000287542,0.005454922,0.0005775477,0.0001673369,0.0002337979,0.0002186616,0.8598297,0.01632774,0.01464953,0.002594286,0.09895599],"study_design_scores_gemma":[0.0001284187,0.0005383437,0.004377194,0.0001259165,0.00007464471,0.0002957596,0.0001071473,0.9396941,0.02030964,0.03210635,0.002186078,0.00005629296],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2548288,0.002633711,0.7292233,0.003176136,0.000190145,0.0002805519,0.000365459,0.001197688,0.00810423],"genre_scores_gemma":[0.8360043,0.00133468,0.1574253,0.0007778743,0.0001450766,0.0003223285,0.0003895479,0.0003643034,0.00323657],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004570615,"threshold_uncertainty_score":0.02417201,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05197575815289519,"score_gpt":0.3078173894272081,"score_spread":0.2558416312743129,"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."}}