{"id":"https://openalex.org/W2797559419","doi":"https://doi.org/10.1162/neco_a_01089","title":"Why Does Large Batch Training Result in Poor Generalization? A Comprehensive Explanation and a Better Strategy from the Viewpoint of Stochastic Optimization","display_name":"Why Does Large Batch Training Result in Poor Generalization? A Comprehensive Explanation and a Better Strategy from the Viewpoint of Stochastic Optimization","publication_year":2018,"publication_date":"2018-04-13","ids":{"openalex":"https://openalex.org/W2797559419","doi":"https://doi.org/10.1162/neco_a_01089","mag":"2797559419","pmid":"https://pubmed.ncbi.nlm.nih.gov/29652590"},"language":"en","primary_location":{"id":"doi:10.1162/neco_a_01089","is_oa":false,"landing_page_url":"https://doi.org/10.1162/neco_a_01089","pdf_url":null,"source":{"id":"https://openalex.org/S207023548","display_name":"Neural Computation","issn_l":"0899-7667","issn":["0899-7667","1530-888X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310315718","host_organization_name":"The MIT Press","host_organization_lineage":["https://openalex.org/P4310315718"],"host_organization_lineage_names":["The MIT Press"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Neural Computation","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","pubmed"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://eprints.lib.hokudai.ac.jp/repo/huscap/all/71558/Tomoumi Takase.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5024516449","display_name":"Tomoumi Takase","orcid":"https://orcid.org/0000-0001-6974-8103"},"institutions":[{"id":"https://openalex.org/I205349734","display_name":"Hokkaido University","ror":"https://ror.org/02e16g702","country_code":"JP","type":"education","lineage":["https://openalex.org/I205349734"]}],"countries":["JP"],"is_corresponding":true,"raw_author_name":"Tomoumi Takase","raw_affiliation_strings":["Graduate School of Information Science and Technology, Hokkaido University, Sapporo, Hokkaido 060-0814, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graduate School of Information Science and Technology, Hokkaido University, Sapporo, Hokkaido 060-0814, Japan","institution_ids":["https://openalex.org/I205349734"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056709028","display_name":"Satoshi Oyama","orcid":"https://orcid.org/0000-0002-8124-3578"},"institutions":[{"id":"https://openalex.org/I205349734","display_name":"Hokkaido University","ror":"https://ror.org/02e16g702","country_code":"JP","type":"education","lineage":["https://openalex.org/I205349734"]}],"countries":["JP"],"is_corresponding":true,"raw_author_name":"Satoshi Oyama","raw_affiliation_strings":["Graduate School of Information Science and Technology, Hokkaido University, Sapporo, Hokkaido 060-0814, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graduate School of Information Science and Technology, Hokkaido University, Sapporo, Hokkaido 060-0814, Japan","institution_ids":["https://openalex.org/I205349734"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5054412358","display_name":"Masahito Kurihara","orcid":"https://orcid.org/0000-0002-1478-1093"},"institutions":[{"id":"https://openalex.org/I205349734","display_name":"Hokkaido University","ror":"https://ror.org/02e16g702","country_code":"JP","type":"education","lineage":["https://openalex.org/I205349734"]}],"countries":["JP"],"is_corresponding":true,"raw_author_name":"Masahito Kurihara","raw_affiliation_strings":["Graduate School of Information Science and Technology, Hokkaido University, Sapporo, Hokkaido 060-0814, Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graduate School of Information Science and Technology, Hokkaido University, Sapporo, Hokkaido 060-0814, Japan","institution_ids":["https://openalex.org/I205349734"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5024516449","https://openalex.org/A5054412358","https://openalex.org/A5056709028"],"corresponding_institution_ids":["https://openalex.org/I205349734"],"apc_list":null,"apc_paid":null,"fwci":0.8459,"has_fulltext":true,"cited_by_count":12,"citation_normalized_percentile":{"value":0.79917642,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":96},"biblio":{"volume":"30","issue":"7","first_page":"2005","last_page":"2023"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12072","display_name":"Machine Learning and Algorithms","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.9995999932289124,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.7476690411567688},{"id":"https://openalex.org/keywords/randomness","display_name":"Randomness","score":0.7051070928573608},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.6248592734336853},{"id":"https://openalex.org/keywords/simulated-annealing","display_name":"Simulated annealing","score":0.6135504245758057},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.6134498119354248},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5744926929473877},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.4784115254878998},{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"Stochastic gradient descent","score":0.4625844657421112},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.4575480818748474},{"id":"https://openalex.org/keywords/gradient-descent","display_name":"Gradient descent","score":0.4505598247051239},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3664480447769165},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.36502599716186523},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.3481740355491638},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.07889127731323242},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.06920391321182251}],"concepts":[{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.7476690411567688},{"id":"https://openalex.org/C125112378","wikidata":"https://www.wikidata.org/wiki/Q176640","display_name":"Randomness","level":2,"score":0.7051070928573608},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.6248592734336853},{"id":"https://openalex.org/C126980161","wikidata":"https://www.wikidata.org/wiki/Q863783","display_name":"Simulated annealing","level":2,"score":0.6135504245758057},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.6134498119354248},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5744926929473877},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.4784115254878998},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.4625844657421112},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.4575480818748474},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.4505598247051239},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3664480447769165},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36502599716186523},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3481740355491638},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.07889127731323242},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.06920391321182251},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C146978453","wikidata":"https://www.wikidata.org/wiki/Q3798668","display_name":"Aerospace engineering","level":1,"score":0.0}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1162/neco_a_01089","is_oa":false,"landing_page_url":"https://doi.org/10.1162/neco_a_01089","pdf_url":null,"source":{"id":"https://openalex.org/S207023548","display_name":"Neural Computation","issn_l":"0899-7667","issn":["0899-7667","1530-888X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310315718","host_organization_name":"The MIT Press","host_organization_lineage":["https://openalex.org/P4310315718"],"host_organization_lineage_names":["The MIT Press"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Neural Computation","raw_type":"journal-article"},{"id":"pmid:29652590","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/29652590","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Neural computation","raw_type":null},{"id":"pmh:oai:irdb.nii.ac.jp:01364:0007185300","is_oa":true,"landing_page_url":"https://hdl.handle.net/2115/71558","pdf_url":"https://eprints.lib.hokudai.ac.jp/repo/huscap/all/71558/Tomoumi Takase.pdf","source":{"id":"https://openalex.org/S7407056385","display_name":"Institutional Repositories DataBase (IRDB)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I184597095","host_organization_name":"National Institute of Informatics","host_organization_lineage":["https://openalex.org/I184597095"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Neural Computation","raw_type":"journal article"},{"id":"pmh:oai:eprints.lib.hokudai.ac.jp:2115/71558","is_oa":false,"landing_page_url":"http://hdl.handle.net/2115/71558","pdf_url":null,"source":{"id":"https://openalex.org/S4306400549","display_name":"Hokkaido University Collection of Scholarly and Academic Papers (Hokkaido University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205349734","host_organization_name":"Hokkaido University","host_organization_lineage":["https://openalex.org/I205349734"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"article"}],"best_oa_location":{"id":"pmh:oai:irdb.nii.ac.jp:01364:0007185300","is_oa":true,"landing_page_url":"https://hdl.handle.net/2115/71558","pdf_url":"https://eprints.lib.hokudai.ac.jp/repo/huscap/all/71558/Tomoumi Takase.pdf","source":{"id":"https://openalex.org/S7407056385","display_name":"Institutional Repositories DataBase (IRDB)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I184597095","host_organization_name":"National Institute of Informatics","host_organization_lineage":["https://openalex.org/I184597095"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Neural Computation","raw_type":"journal article"},"sustainable_development_goals":[{"score":0.6800000071525574,"display_name":"No poverty","id":"https://metadata.un.org/sdg/1"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2797559419.pdf","grobid_xml":"https://content.openalex.org/works/W2797559419.grobid-xml"},"referenced_works_count":27,"referenced_works":["https://openalex.org/W104184427","https://openalex.org/W1533861849","https://openalex.org/W1983624953","https://openalex.org/W2024060531","https://openalex.org/W2024860775","https://openalex.org/W2047422346","https://openalex.org/W2084792706","https://openalex.org/W2095705004","https://openalex.org/W2107438106","https://openalex.org/W2112796928","https://openalex.org/W2126554879","https://openalex.org/W2146502635","https://openalex.org/W2154061444","https://openalex.org/W2156387975","https://openalex.org/W2253093986","https://openalex.org/W2559655401","https://openalex.org/W2593380010","https://openalex.org/W2604272474","https://openalex.org/W2607662938","https://openalex.org/W2766164908","https://openalex.org/W2913668833","https://openalex.org/W2949117887","https://openalex.org/W2950928354","https://openalex.org/W2963959597","https://openalex.org/W2964121744","https://openalex.org/W4238404964","https://openalex.org/W4300537377"],"related_works":["https://openalex.org/W2895097035","https://openalex.org/W4206903459","https://openalex.org/W2754816816","https://openalex.org/W4366280654","https://openalex.org/W3160167280","https://openalex.org/W3020853991","https://openalex.org/W4362706668","https://openalex.org/W4231621013","https://openalex.org/W3171021120","https://openalex.org/W3008318776"],"abstract_inverted_index":{"We":[0,69],"present":[1],"a":[2,23],"comprehensive":[3],"framework":[4,48],"of":[5,46],"search":[6,22],"methods,":[7],"such":[8],"as":[9],"simulated":[10],"annealing":[11],"and":[12],"batch":[13,65,74],"training,":[14],"for":[15],"solving":[16],"nonconvex":[17],"optimization":[18],"problems.":[19],"These":[20],"methods":[21],"wider":[24],"range":[25],"by":[26],"gradually":[27,63],"decreasing":[28],"the":[29,33,44],"randomness":[30],"added":[31],"to":[32,53],"standard":[34],"gradient":[35],"descent":[36],"method.":[37],"The":[38],"formulation":[39],"that":[40,62],"we":[41],"define":[42],"on":[43],"basis":[45],"this":[47],"can":[49],"be":[50],"directly":[51],"applied":[52],"neural":[54],"network":[55],"training.":[56,68],"This":[57],"produces":[58],"an":[59],"effective":[60],"approach":[61],"increases":[64],"size":[66],"during":[67],"also":[70],"explain":[71],"why":[72],"large":[73],"training":[75],"degrades":[76],"generalization":[77],"performance,":[78],"which":[79],"previous":[80],"studies":[81],"have":[82],"not":[83],"clarified.":[84]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":2},{"year":2022,"cited_by_count":2},{"year":2021,"cited_by_count":1},{"year":2020,"cited_by_count":2},{"year":2019,"cited_by_count":2}],"updated_date":"2026-05-21T06:26:12.895304","created_date":"2025-10-10T00:00:00"}