{"id":"https://openalex.org/W4304477701","doi":"https://doi.org/10.1109/tpami.2022.3213654","title":"An Efficient Fisher Matrix Approximation Method for Large-Scale Neural Network Optimization","display_name":"An Efficient Fisher Matrix Approximation Method for Large-Scale Neural Network Optimization","publication_year":2022,"publication_date":"2022-01-01","ids":{"openalex":"https://openalex.org/W4304477701","doi":"https://doi.org/10.1109/tpami.2022.3213654","pmid":"https://pubmed.ncbi.nlm.nih.gov/36219666"},"language":"en","primary_location":{"id":"doi:10.1109/tpami.2022.3213654","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpami.2022.3213654","pdf_url":null,"source":{"id":"https://openalex.org/S199944782","display_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","issn_l":"0162-8828","issn":["0162-8828","1939-3539","2160-9292"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320439","host_organization_name":"IEEE Computer Society","host_organization_lineage":["https://openalex.org/P4310320439","https://openalex.org/P4310319808"],"host_organization_lineage_names":["IEEE Computer Society","Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","pubmed"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101802213","display_name":"Minghan Yang","orcid":"https://orcid.org/0000-0002-6501-7132"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Minghan Yang","raw_affiliation_strings":["DAMO Academy, Alibaba Group, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"DAMO Academy, Alibaba Group, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100341937","display_name":"Dong Xu","orcid":"https://orcid.org/0000-0002-9451-1754"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]},{"id":"https://openalex.org/I4210133846","display_name":"Peking University International Hospital","ror":"https://ror.org/03jxhcr96","country_code":"CN","type":"healthcare","lineage":["https://openalex.org/I4210133846"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dong Xu","raw_affiliation_strings":["Beijing International Center for Mathematical Research, Peking University, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing International Center for Mathematical Research, Peking University, China","institution_ids":["https://openalex.org/I20231570","https://openalex.org/I4210133846"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027905775","display_name":"Qiwen Cui","orcid":null},"institutions":[{"id":"https://openalex.org/I201448701","display_name":"University of Washington","ror":"https://ror.org/00cvxb145","country_code":"US","type":"education","lineage":["https://openalex.org/I201448701"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Qiwen Cui","raw_affiliation_strings":["Paul G. Allen School of Computer Science & Engineering, University of Washington, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Paul G. Allen School of Computer Science & Engineering, University of Washington, USA","institution_ids":["https://openalex.org/I201448701"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006127137","display_name":"Zaiwen Wen","orcid":"https://orcid.org/0000-0003-1762-0671"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zaiwen Wen","raw_affiliation_strings":["Beijing International Center for Mathematical Research, College of Engineering and Center for Machine Learning Research, Peking University, China"],"raw_orcid":"https://orcid.org/0000-0003-1762-0671","affiliations":[{"raw_affiliation_string":"Beijing International Center for Mathematical Research, College of Engineering and Center for Machine Learning Research, Peking University, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5007885231","display_name":"Pengxiang Xu","orcid":"https://orcid.org/0000-0001-7630-2919"},"institutions":[{"id":"https://openalex.org/I4210136793","display_name":"Peng Cheng Laboratory","ror":"https://ror.org/03qdqbt06","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210136793"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Pengxiang Xu","raw_affiliation_strings":["Peng Cheng Laboratory, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peng Cheng Laboratory, China","institution_ids":["https://openalex.org/I4210136793"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5101802213"],"corresponding_institution_ids":["https://openalex.org/I45928872"],"apc_list":null,"apc_paid":null,"fwci":0.6937,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.75538687,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":98},"biblio":{"volume":"45","issue":"5","first_page":"1","last_page":"13"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12676","display_name":"Machine Learning and ELM","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"}},"topics":[{"id":"https://openalex.org/T12676","display_name":"Machine Learning and ELM","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"}},{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.9983000159263611,"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/T10057","display_name":"Face and Expression Recognition","score":0.9970999956130981,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/computer-science","display_name":"Computer science","score":0.5977181792259216},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5647301077842712},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5190642476081848},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.4305216670036316},{"id":"https://openalex.org/keywords/matrix","display_name":"Matrix (chemical analysis)","score":0.413652241230011},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.39656558632850647},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3626400828361511},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3605387508869171},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.30199289321899414},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.09547439217567444}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5977181792259216},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5647301077842712},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5190642476081848},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.4305216670036316},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.413652241230011},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.39656558632850647},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3626400828361511},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3605387508869171},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.30199289321899414},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.09547439217567444},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C192562407","wikidata":"https://www.wikidata.org/wiki/Q228736","display_name":"Materials science","level":0,"score":0.0},{"id":"https://openalex.org/C159985019","wikidata":"https://www.wikidata.org/wiki/Q181790","display_name":"Composite material","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/tpami.2022.3213654","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpami.2022.3213654","pdf_url":null,"source":{"id":"https://openalex.org/S199944782","display_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","issn_l":"0162-8828","issn":["0162-8828","1939-3539","2160-9292"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320439","host_organization_name":"IEEE Computer Society","host_organization_lineage":["https://openalex.org/P4310320439","https://openalex.org/P4310319808"],"host_organization_lineage_names":["IEEE Computer Society","Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","raw_type":"journal-article"},{"id":"pmid:36219666","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/36219666","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":"IEEE transactions on pattern analysis and machine intelligence","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":41,"referenced_works":["https://openalex.org/W1480347379","https://openalex.org/W1992985800","https://openalex.org/W1994616650","https://openalex.org/W2080635178","https://openalex.org/W2108598243","https://openalex.org/W2129160848","https://openalex.org/W2146502635","https://openalex.org/W2194775991","https://openalex.org/W2622263826","https://openalex.org/W2726670313","https://openalex.org/W2757910899","https://openalex.org/W2810281214","https://openalex.org/W2899283552","https://openalex.org/W2900789157","https://openalex.org/W2963060476","https://openalex.org/W2963941964","https://openalex.org/W2964303576","https://openalex.org/W2979245724","https://openalex.org/W2988070836","https://openalex.org/W3034315405","https://openalex.org/W3035182906","https://openalex.org/W3086499488","https://openalex.org/W3133629262","https://openalex.org/W3185298382","https://openalex.org/W4289203231","https://openalex.org/W6628707872","https://openalex.org/W6631190155","https://openalex.org/W6681435938","https://openalex.org/W6683107984","https://openalex.org/W6683993359","https://openalex.org/W6684497637","https://openalex.org/W6713134421","https://openalex.org/W6721263324","https://openalex.org/W6739622702","https://openalex.org/W6744513255","https://openalex.org/W6748982233","https://openalex.org/W6766978945","https://openalex.org/W6769062451","https://openalex.org/W6770934496","https://openalex.org/W6779861650","https://openalex.org/W6791581255"],"related_works":["https://openalex.org/W2391251536","https://openalex.org/W2362198218","https://openalex.org/W1982750869","https://openalex.org/W2019521278","https://openalex.org/W1984922432","https://openalex.org/W2113077220","https://openalex.org/W2375008505","https://openalex.org/W2350679292","https://openalex.org/W2086348228","https://openalex.org/W4390653028"],"abstract_inverted_index":{"Although":[0],"the":[1,4,26,29,52,55,71,78,89,94,104,109,114,168],"shapes":[2],"of":[3,28,73,77,88],"parameters":[5,53],"are":[6,135],"not":[7],"crucial":[8],"for":[9],"designing":[10],"first-order":[11,124],"optimization":[12],"methods":[13],"in":[14,54,93],"large":[15],"scale":[16],"empirical":[17],"risk":[18],"minimization":[19],"problems,":[20],"they":[21],"have":[22],"important":[23],"impact":[24],"on":[25,51,142],"size":[27,76],"matrix":[30,57,67,79,92,105],"to":[31,80,167],"be":[32,81,118],"inverted":[33,82],"when":[34],"developing":[35],"second-order":[36,48],"type":[37],"methods.":[38,125,170],"In":[39],"this":[40],"article,":[41],"we":[42],"propose":[43],"an":[44],"efficient":[45],"and":[46,62,108,120,129,152],"novel":[47],"method":[49,163],"based":[50],"real":[56,95],"space":[58,97],"[Formula:":[59,98],"see":[60,99],"text]":[61],"a":[63,130],"matrix-product":[64],"approximate":[65],"Fisher":[66,90],"(MatFisher)":[68],"by":[69,102],"using":[70],"products":[72],"gradients.":[74],"The":[75],"is":[83,121,164],"much":[84],"smaller":[85],"than":[86],"that":[87,161],"information":[91],"vector":[96],"text].":[100],"Moreover,":[101],"utilizing":[103],"delayed":[106],"update":[107],"block":[110],"diagonal":[111],"approximation":[112],"techniques,":[113],"computational":[115],"cost":[116],"can":[117],"controlled":[119],"comparable":[122],"with":[123,145,150,158],"A":[126],"global":[127],"convergence":[128,133],"superlinear":[131],"local":[132],"analysis":[134],"established":[136],"under":[137],"mild":[138],"conditions.":[139],"Numerical":[140],"results":[141],"image":[143],"classification":[144],"ResNet50,":[146],"quantum":[147],"chemistry":[148],"modeling":[149],"SchNet,":[151],"data-driven":[153],"partial":[154],"differential":[155],"equations":[156],"solution":[157],"PINN":[159],"illustrate":[160],"our":[162],"quite":[165],"competitive":[166],"state-of-the-art":[169]},"counts_by_year":[{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}