{"id":"https://openalex.org/W3195437175","doi":null,"title":"FedPara: Low-rank Hadamard Product Parameterization for Efficient Federated Learning","display_name":"FedPara: Low-rank Hadamard Product Parameterization for Efficient Federated Learning","publication_year":2021,"publication_date":"2021-08-13","ids":{"openalex":"https://openalex.org/W3195437175","mag":"3195437175"},"language":"en","primary_location":{"id":"mag:3195437175","doi":null,"is_oa":true,"landing_page_url":"http://export.arxiv.org/pdf/2108.06098","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"http://export.arxiv.org/pdf/2108.06098","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5090138474","display_name":"Nam Hyeon-Woo","orcid":"https://orcid.org/0000-0001-9543-3770","observed_orcids":["https://orcid.org/0000-0001-9543-3770"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hyeon-Woo Nam","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0001-9543-3770","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5007559554","display_name":"Moon Ye-Bin","orcid":"https://orcid.org/0000-0002-0390-6567","observed_orcids":["https://orcid.org/0000-0002-0390-6567"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Moon Ye-Bin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5078114111","display_name":"Tae-Hyun Oh","orcid":"https://orcid.org/0000-0003-0468-1571","observed_orcids":["https://orcid.org/0000-0003-0468-1571"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tae-Hyun Oh","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0003-0468-1571","affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9994999766349792,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9994999766349792,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.00019999999494757503,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":9.999999747378752e-05,"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/federated-learning","display_name":"federated learning","score":0.9679999947547913},{"id":"https://openalex.org/keywords/communication-efficiency","display_name":"communication efficiency","score":0.8840000033378601},{"id":"https://openalex.org/keywords/hadamard-product","display_name":"Hadamard product","score":0.996999979019165},{"id":"https://openalex.org/keywords/reparameterization","display_name":"reparameterization","score":0.800000011920929},{"id":"https://openalex.org/keywords/personalized-federated-learning","display_name":"personalized federated learning","score":0.9789999723434448},{"id":"https://openalex.org/keywords/model-compression","display_name":"model compression","score":0.4690000116825104}],"concepts":[{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.7937591075897217},{"id":"https://openalex.org/C71901391","wikidata":"https://www.wikidata.org/wiki/Q7126699","display_name":"Upload","level":2,"score":0.7225828170776367},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6761203408241272},{"id":"https://openalex.org/C60292330","wikidata":"https://www.wikidata.org/wiki/Q1014065","display_name":"Hadamard transform","level":2,"score":0.5997106432914734},{"id":"https://openalex.org/C90673727","wikidata":"https://www.wikidata.org/wiki/Q901718","display_name":"Product (mathematics)","level":2,"score":0.5117999911308289},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2638534903526306},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.05881643295288086},{"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/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"mag:3195437175","doi":null,"is_oa":true,"landing_page_url":"http://export.arxiv.org/pdf/2108.06098","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null}],"best_oa_location":{"id":"mag:3195437175","doi":null,"is_oa":true,"landing_page_url":"http://export.arxiv.org/pdf/2108.06098","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":6,"referenced_works":["https://openalex.org/W2535838896","https://openalex.org/W2773126525","https://openalex.org/W2895171208","https://openalex.org/W2990789643","https://openalex.org/W2995022099","https://openalex.org/W3043638540"],"related_works":["https://openalex.org/W3209119160","https://openalex.org/W3088296980","https://openalex.org/W7164497647","https://openalex.org/W4406458349","https://openalex.org/W7160204372","https://openalex.org/W7106237959","https://openalex.org/W7221796695","https://openalex.org/W4403617805","https://openalex.org/W3216443854","https://openalex.org/W4400034433"],"abstract_inverted_index":{"To":[0],"overcome":[1],"the":[2,23,33,38,58,63,74,87,95,100],"burdens":[3],"on":[4],"frequent":[5],"model":[6],"uploads":[7],"and":[8,110,138],"downloads":[9],"during":[10],"federated":[11],"learning":[12],"(FL),":[13],"we":[14],"propose":[15],"a":[16,54,128],"communication-efficient":[17],"re-parameterization,":[18],"FedPara.":[19],"Our":[20],"method":[21,43,109,117,126],"re-parameterizes":[22],"model's":[24],"layers":[25],"using":[26],"low-rank":[27,40,48,59,97],"matrices":[28],"or":[29],"tensors":[30],"followed":[31],"by":[32,94,106],"Hadamard":[34],"product.":[35],"Different":[36],"from":[37],"conventional":[39],"parameterization,":[41],"our":[42,51,108,116,125],"is":[44,91,118],"not":[45,92],"limited":[46],"to":[47,73,80,127],"constraints.":[49],"Thereby,":[50],"FedPara":[52],"has":[53],"larger":[55],"capacity":[56],"than":[57,86,152],"one,":[60],"even":[61],"with":[62,120,150],"same":[64],"number":[65],"of":[66],"parameters.":[67,156],"It":[68],"can":[69,102],"achieve":[70],"comparable":[71],"performance":[72],"original":[75,88],"models":[76],"while":[77],"requiring":[78],"2.8":[79],"10.1":[81],"times":[82,154],"lower":[83],"communication":[84],"costs":[85],"models,":[89],"which":[90,133],"achievable":[93],"traditional":[96],"parameterization.":[98],"Moreover,":[99],"efficiency":[101],"be":[103],"further":[104],"improved":[105],"combining":[107],"other":[111],"efficient":[112],"FL":[113,130,148],"techniques":[114],"because":[115],"compatible":[119],"others.":[121],"We":[122,141],"also":[123],"extend":[124],"personalized":[129,147],"application,":[131],"pFedPara,":[132],"separates":[134],"parameters":[135],"into":[136],"global":[137],"local":[139],"ones.":[140],"show":[142],"that":[143],"pFedPara":[144],"outperforms":[145],"competing":[146],"methods":[149],"more":[151],"three":[153],"fewer":[155]},"counts_by_year":[{"year":2023,"cited_by_count":1},{"year":2021,"cited_by_count":1}],"updated_date":"2026-10-13T10:08:12.033832","created_date":"2021-08-30T00:00:00"}