{"id":"W4319339631","doi":"10.36227/techrxiv.22006742.v1","title":"Improving Image Recognition Accuracy in Neural Networks Using Fractional Natural Gradient Descent","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Image Processing Techniques and Applications","field":"Engineering","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Ministry of Science and Higher Education of the Russian Federation; Russian Science Foundation; Centre de Recherches Mathématiques","keywords":"MNIST database; Convexity; Gradient descent; Mathematics; Curvature; Rate of convergence; Convergence (economics); Artificial neural network; Momentum (technical analysis); Image (mathematics); Order (exchange); Applied mathematics; Function (biology); Algorithm; Computer science; Artificial intelligence; Geometry","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0001756617,0.0002926085,0.000224851,0.000236703,0.00009486234,0.000230733,0.000227542,0.0002343145,0.00002800122],"category_scores_gemma":[0.00005636538,0.0003164879,0.0001089912,0.0002582626,0.00003044261,0.0002581421,0.0003029845,0.001338956,0.00001287485],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003840406,"about_ca_system_score_gemma":0.00003472428,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000421541,"about_ca_topic_score_gemma":0.00006490626,"domain_scores_codex":[0.9986402,0.00001965499,0.0004216535,0.0003950714,0.000152282,0.0003711424],"domain_scores_gemma":[0.9993905,0.00007623184,0.0001117705,0.0002716971,0.0000945432,0.00005526783],"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":[0.00001767132,0.00009652411,0.0002381665,0.0008386999,0.00005245619,0.00003985599,0.0001184689,0.7111086,0.01269093,0.0000310967,0.001992756,0.2727748],"study_design_scores_gemma":[0.00009611172,0.000003049193,0.0004991978,0.0001900173,0.00002077596,0.00001272848,0.00002122619,0.9959669,0.001602695,0.001202174,0.00002843362,0.0003566677],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05514826,0.0004361017,0.9402736,0.0001173407,0.00115939,0.0005832224,0.00002570927,0.002121767,0.0001345691],"genre_scores_gemma":[0.8965088,0.0001411391,0.1021776,0.00005353085,0.000361325,0.0002375929,0.0003807838,0.0001039646,0.00003525692],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8413606,"threshold_uncertainty_score":0.9999287,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04066706910852975,"score_gpt":0.2923875884603741,"score_spread":0.2517205193518444,"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."}}