{"id":"W2280728065","doi":"","title":"Stage-wise Training: An Improved Feature Learning Strategy for Deep Models","year":2015,"lang":"en","type":"article","venue":"Neural Information Processing Systems","topic":"Image and Signal Denoising Methods","field":"Computer Science","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Artificial intelligence; Regularization (linguistics); Curse of dimensionality; Machine learning; Artificial neural network; Decoupling (probability); Deep learning; Feature (linguistics); Context (archaeology); Process (computing); Generalization; Feature extraction; Feature learning; Mathematics; Engineering","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":[0.001263975,0.001237136,0.0009397142,0.0004633181,0.0004064733,0.0005850799,0.002461368,0.001354467,0.002967063],"category_scores_gemma":[0.003069004,0.0008333494,0.0009318423,0.0005834359,0.0006038182,0.00182178,0.001434081,0.002509183,0.001059328],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005515408,"about_ca_system_score_gemma":0.0009595357,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003250159,"about_ca_topic_score_gemma":0.006180861,"domain_scores_codex":[0.999621,0.000114942,0.00002551724,0.00008267997,0.0001014215,0.00005445833],"domain_scores_gemma":[0.9992525,0.0002930835,0.00005380382,0.0001990416,0.0001645384,0.00003696532],"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":[0.0002369273,0.0002080273,0.00146383,0.0001348314,0.00009508829,0.0001616041,0.0001356274,0.380682,0.03077076,0.02402978,0.005908151,0.5561733],"study_design_scores_gemma":[0.000007699313,0.00003941467,0.00007219146,0.000004349751,0.000007511583,0.00001871548,0.000002988446,0.9925685,0.003437903,0.003206261,0.0006289608,0.000005473691],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.003905752,0.00008673722,0.9946779,0.00007441272,0.00001494294,0.00004100554,0.00003365312,0.0006676272,0.0004980401],"genre_scores_gemma":[0.2353538,0.0002128844,0.7589182,0.0002246516,0.00006282382,0.0003030944,0.0003814881,0.0003500816,0.004192985],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.003250159,"threshold_uncertainty_score":0.009925783,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1193691106511897,"score_gpt":0.3293162928280917,"score_spread":0.2099471821769019,"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."}}