{"id":"https://openalex.org/W3093034895","doi":null,"title":"FedAT: A Communication-Efficient Federated Learning Method with Asynchronous Tiers under Non-IID Data.","display_name":"FedAT: A Communication-Efficient Federated Learning Method with Asynchronous Tiers under Non-IID Data.","publication_year":2020,"publication_date":"2020-10-12","ids":{"openalex":"https://openalex.org/W3093034895","mag":"3093034895"},"language":"en","primary_location":{"id":"mag:3093034895","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2010.05958","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":"https://arxiv.org/abs/2010.05958","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5081043925","display_name":"Zheng Chai","orcid":"https://orcid.org/0000-0001-9823-2990","observed_orcids":["https://orcid.org/0000-0001-9823-2990","https://orcid.org/0009-0001-2092-8426","https://orcid.org/0009-0003-5340-0673"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng Chai","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0009-0003-5340-0673","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101451077","display_name":"Yujing Chen","orcid":"https://orcid.org/0009-0000-6921-964X","observed_orcids":["https://orcid.org/0009-0000-6921-964X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yujing Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054490777","display_name":"Liang Zhao","orcid":"https://orcid.org/0000-0001-6706-2271","observed_orcids":["https://orcid.org/0000-0001-6706-2271"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liang Zhao","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0001-6706-2271","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5079187166","display_name":"Yue Cheng","orcid":"https://orcid.org/0000-0003-1695-4864","observed_orcids":["https://orcid.org/0000-0003-1695-4864"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yue Cheng","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0003-1695-4864","affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5006581225","display_name":"Huzefa Rangwala","orcid":"https://orcid.org/0000-0003-0435-0035","observed_orcids":["https://orcid.org/0000-0003-0435-0035"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huzefa Rangwala","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0003-0435-0035","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":66,"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":0.0003000000142492354,"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":0.0,"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":0.9900000095367432},{"id":"https://openalex.org/keywords/non-iid-data","display_name":"non-IID data","score":0.8349999785423279},{"id":"https://openalex.org/keywords/straggler-problem","display_name":"straggler problem","score":0.9259999990463257},{"id":"https://openalex.org/keywords/communication-bottleneck","display_name":"communication bottleneck","score":0.7630000114440918},{"id":"https://openalex.org/keywords/weighted-aggregation","display_name":"weighted aggregation","score":0.6650000214576721},{"id":"https://openalex.org/keywords/communication-efficiency","display_name":"communication efficiency","score":0.7089999914169312},{"id":"https://openalex.org/keywords/convergence-rate","display_name":"convergence rate","score":0.7039999961853027},{"id":"https://openalex.org/keywords/client-heterogeneity","display_name":"client heterogeneity","score":0.5199999809265137}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8541700839996338},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.841096043586731},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.8299402594566345},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5641746520996094},{"id":"https://openalex.org/C2992525071","wikidata":"https://www.wikidata.org/wiki/Q50818671","display_name":"Federated learning","level":2,"score":0.484570175409317},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.4215627610683441},{"id":"https://openalex.org/C2780945871","wikidata":"https://www.wikidata.org/wiki/Q194274","display_name":"Backup","level":2,"score":0.417675256729126},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.13555526733398438},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.11342629790306091}],"mesh":[],"locations_count":1,"locations":[{"id":"mag:3093034895","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2010.05958","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:3093034895","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2010.05958","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":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":16,"referenced_works":["https://openalex.org/W2016321198","https://openalex.org/W2113839990","https://openalex.org/W2436219157","https://openalex.org/W2460676686","https://openalex.org/W2541884796","https://openalex.org/W2750384547","https://openalex.org/W2804294608","https://openalex.org/W2804978897","https://openalex.org/W2963300197","https://openalex.org/W2975043678","https://openalex.org/W2995653155","https://openalex.org/W2995735216","https://openalex.org/W2998045710","https://openalex.org/W3037871107","https://openalex.org/W3043303805","https://openalex.org/W3118608800"],"related_works":["https://openalex.org/W3197301111","https://openalex.org/W3208693455","https://openalex.org/W4360980584","https://openalex.org/W4312839427","https://openalex.org/W4394712191","https://openalex.org/W7167353995","https://openalex.org/W4399766962","https://openalex.org/W4225874127","https://openalex.org/W3216265618","https://openalex.org/W4402568979"],"abstract_inverted_index":{"Federated":[0,143],"learning":[1,21,144],"(FL)":[2],"involves":[3],"training":[4,14,157,168,195],"a":[5,62,72,116,141,185],"model":[6,26,29,98],"over":[7],"massive":[8],"distributed":[9],"devices,":[10],"while":[11,120],"keeping":[12],"the":[13,45,77,91,106,110,164,173,194,202,215,223,232],"data":[15,50],"localized.":[16],"This":[17],"form":[18],"of":[19,65,90],"collaborative":[20],"exposes":[22],"new":[23,38],"tradeoffs":[24],"among":[25],"convergence":[27,178],"speed,":[28],"accuracy,":[30],"balance":[31,193],"across":[32],"clients,":[33],"and":[34,53,57,75,158,166,180,192,204,230],"communication":[35,59,118,216,233],"cost,":[36],"with":[37,130,146,176],"challenges":[39],"including:":[40],"(1)":[41],"straggler":[42,107,174],"problem,":[43],"where":[44,61],"clients":[46,66],"lag":[47],"due":[48],"to":[49,71,104,190,228,237,240],"or":[51,100],"(computing":[52],"network)":[54],"resource":[55],"heterogeneity,":[56],"(2)":[58],"bottleneck,":[60,119],"large":[63],"number":[64],"communicate":[67],"their":[68],"local":[69],"updates":[70],"central":[73],"server":[74],"bottleneck":[76],"server.":[78],"Many":[79],"existing":[80],"FL":[81,242],"methods":[82,112],"focus":[83],"on":[84],"optimizing":[85],"along":[86],"only":[87],"one":[88],"dimension":[89],"tradeoff":[92],"space.":[93],"Existing":[94],"solutions":[95],"use":[96],"asynchronous":[97,111,159,167],"updating":[99],"tiering-based":[101],"synchronous":[102,155,165],"mechanisms":[103],"tackle":[105],"problem.":[108],"However,":[109],"can":[113],"easily":[114],"create":[115],"network":[117],"tiering":[121,126],"may":[122],"introduce":[123],"biases":[124],"as":[125],"favors":[127],"faster":[128],"tiers":[129],"shorter":[131],"response":[132],"latencies.":[133],"To":[134],"address":[135],"these":[136],"issues,":[137],"we":[138],"present":[139],"FedAT,":[140],"novel":[142],"method":[145],"Asynchronous":[147],"Tiers":[148],"under":[149],"Non-i.i.d.":[150],"data.":[151],"FedAT":[152,171,183,200,221],"synergistically":[153],"combines":[154],"intra-tier":[156],"cross-tier":[160],"training.":[161],"By":[162],"bridging":[163],"through":[169],"tiering,":[170],"minimizes":[172],"effect":[175],"improved":[177],"speed":[179],"test":[181],"accuracy.":[182],"uses":[184],"straggler-aware,":[186],"weighted":[187],"aggregation":[188],"heuristic":[189],"steer":[191],"for":[196],"further":[197],"accuracy":[198],"improvement.":[199],"compresses":[201],"uplink":[203],"downlink":[205],"communications":[206],"using":[207],"an":[208],"efficient,":[209],"polyline-encoding-based":[210],"compression":[211],"algorithm,":[212],"therefore":[213],"minimizing":[214],"cost.":[217],"Results":[218],"show":[219],"that":[220],"improves":[222],"prediction":[224],"performance":[225],"by":[226,235],"up":[227,236],"21.09%,":[229],"reduces":[231],"cost":[234],"8.5x,":[238],"compared":[239],"state-of-the-art":[241],"methods.":[243]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":11},{"year":2023,"cited_by_count":23},{"year":2022,"cited_by_count":20},{"year":2021,"cited_by_count":10}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2020-10-22T00:00:00"}