Robust Federated Learning With Noisy Labeled Data Through Loss Function Correction

Federated learning (FL) is a communication-efficient machine learning paradigm to leverage distributed data at the network edge. Nevertheless, FL usually fails to train a high-quality model from the networks, where the edge nodes collect noisy labeled data. To tackle this challenge, this paper focus...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:IEEE transactions on network science and engineering Ročník 10; číslo 3; s. 1 - 11
Hlavní autori: Chen, Li, Ang, Fan, Chen, Yunfei
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Piscataway IEEE 01.05.2023
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Predmet:
ISSN:2327-4697, 2334-329X
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract Federated learning (FL) is a communication-efficient machine learning paradigm to leverage distributed data at the network edge. Nevertheless, FL usually fails to train a high-quality model from the networks, where the edge nodes collect noisy labeled data. To tackle this challenge, this paper focuses on developing an innovative robust FL. We consider two kinds of networks with different data distribution. Firstly, we design a reweighted FL under a full-data network, where all edge nodes are equipped with both numerous noisy labeled dataset and small clean dataset. The key idea is that edge devices learn to assign the local weights of loss functions in noisy labeled dataset, and cooperate with central server to update global weights. Secondly, we consider a part-data network where some edge nodes exclude clean dataset, and can not compute the weights locally. The broadcasting of the global weights is added to help those edge nodes without clean dataset to reweight their noisy loss functions. Both designs have a convergence rate of <inline-formula><tex-math notation="LaTeX">\mathcal {O}(1/T^{2})</tex-math></inline-formula>. Simulation results illustrate that the both proposed training processes improve the prediction accuracy due to the proper weights assignments of noisy loss function.
AbstractList Federated learning (FL) is a communication-efficient machine learning paradigm to leverage distributed data at the network edge. Nevertheless, FL usually fails to train a high-quality model from the networks, where the edge nodes collect noisy labeled data. To tackle this challenge, this paper focuses on developing an innovative robust FL. We consider two kinds of networks with different data distribution. Firstly, we design a reweighted FL under a full-data network, where all edge nodes are equipped with both numerous noisy labeled dataset and small clean dataset. The key idea is that edge devices learn to assign the local weights of loss functions in noisy labeled dataset, and cooperate with central server to update global weights. Secondly, we consider a part-data network where some edge nodes exclude clean dataset, and can not compute the weights locally. The broadcasting of the global weights is added to help those edge nodes without clean dataset to reweight their noisy loss functions. Both designs have a convergence rate of <inline-formula><tex-math notation="LaTeX">\mathcal {O}(1/T^{2})</tex-math></inline-formula>. Simulation results illustrate that the both proposed training processes improve the prediction accuracy due to the proper weights assignments of noisy loss function.
Federated learning (FL) is a communication-efficient machine learning paradigm to leverage distributed data at the network edge. Nevertheless, FL usually fails to train a high-quality model from the networks, where the edge nodes collect noisy labeled data. To tackle this challenge, this paper focuses on developing an innovative robust FL. We consider two kinds of networks with different data distribution. Firstly, we design a reweighted FL under a full-data network, where all edge nodes are equipped with both numerous noisy labeled dataset and small clean dataset. The key idea is that edge devices learn to assign the local weights of loss functions in noisy labeled dataset, and cooperate with central server to update global weights. Secondly, we consider a part-data network where some edge nodes exclude clean dataset, and can not compute the weights locally. The broadcasting of the global weights is added to help those edge nodes without clean dataset to reweight their noisy loss functions. Both designs have a convergence rate of [Formula Omitted]. Simulation results illustrate that the both proposed training processes improve the prediction accuracy due to the proper weights assignments of noisy loss function.
Author Chen, Yunfei
Chen, Li
Ang, Fan
Author_xml – sequence: 1
  givenname: Li
  orcidid: 0000-0002-1754-0607
  surname: Chen
  fullname: Chen, Li
  organization: Department of Electronic Engineering and Information Science, University of Science and Technology of China, China
– sequence: 2
  givenname: Fan
  orcidid: 0000-0003-2569-745X
  surname: Ang
  fullname: Ang, Fan
  organization: Department of Electronic Engineering and Information Science, University of Science and Technology of China, China
– sequence: 3
  givenname: Yunfei
  orcidid: 0000-0001-8083-1805
  surname: Chen
  fullname: Chen, Yunfei
  organization: School of Engineering, University of Warwick, Coventry, U.K
BookMark eNp9kDFPwzAQhS1UJErpD0AslphT7HMbxyMqLSBFRSpFsEV2cmlTlbg4ztB_j0MrBgamO-neu_vuXZJebWsk5JqzEedM3a0Wr7MRMICRAJCQyDPSByHGkQD10et6kNE4VvKCDJtmyxjjkMRCiD5ZLq1pG0_nWKDTHguaonZ1Va_pe-U3dGGr5kBTbXAXZg_aa7raONuuNzS1TUPnbZ37ytZ0ap3Dn_aKnJd61-DwVAfkbT5bTZ-i9OXxeXqfRjko4aPAyRNTIhquUCupJmWSawnhIywDXKAsE8bQFNpIJaGQzMRg4pLDGIUWYkBuj3v3zn612Phsa1tXh5MZJCwGHiuVBJU8qnIXeB2WWV553XF6p6tdxlnWZZh1GWZdhtkpw-Dkf5x7V31qd_jXc3P0VIj4q1dKCjHh4htIG34M
CODEN ITNSD5
CitedBy_id crossref_primary_10_1109_TBDATA_2025_3527202
crossref_primary_10_1080_23742917_2024_2448355
crossref_primary_10_1111_exsy_70096
crossref_primary_10_1109_TNSE_2023_3320123
crossref_primary_10_1016_j_mex_2025_103408
crossref_primary_10_1145_3727643
crossref_primary_10_1109_TSIPN_2025_3572292
Cites_doi 10.1561/9781680837896
10.1109/ICCV.2015.168
10.1109/TWC.2020.2974748
10.1109/TWC.2021.3099505
10.1109/CVPR.2016.90
10.1109/JIOT.2020.3023126
10.1109/TNSE.2020.3014385
10.1109/MWC.001.1900119
10.1109/TNNLS.2022.3152527
10.1109/CVPR.2015.7298885
10.1109/JIOT.2020.3027980
10.1109/MCOM.2019.1900271
10.1007/s10915-018-0757-z
10.1109/TII.2019.2938861
10.1145/3298981
10.1109/WCNC45663.2020.9120713
10.1109/JSAC.2019.2904348
10.1109/TAC.2014.2308612
10.1609/aaai.v31i1.10894
10.1109/TCOMM.2020.2979149
10.1109/TNSE.2020.3016035
10.1109/TGRS.2019.2961141
10.1109/TCOMM.2021.3081746
ContentType Journal Article
Copyright Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023
Copyright_xml – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023
DBID 97E
RIA
RIE
AAYXX
CITATION
7SC
8FD
JQ2
L7M
L~C
L~D
DOI 10.1109/TNSE.2022.3227287
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005–Present
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE Electronic Library (IEL)
CrossRef
Computer and Information Systems Abstracts
Technology Research Database
ProQuest Computer Science Collection
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts – Academic
Computer and Information Systems Abstracts Professional
DatabaseTitle CrossRef
Computer and Information Systems Abstracts
Technology Research Database
Computer and Information Systems Abstracts – Academic
Advanced Technologies Database with Aerospace
ProQuest Computer Science Collection
Computer and Information Systems Abstracts Professional
DatabaseTitleList
Computer and Information Systems Abstracts
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISSN 2334-329X
EndPage 11
ExternalDocumentID 10_1109_TNSE_2022_3227287
9973351
Genre orig-research
GrantInformation_xml – fundername: Fundamental Research Funds for the Central Universities
  grantid: WK3500000007
– fundername: National Key Research and Development Program of China
  grantid: 2018YFA0701603
– fundername: National Natural Science Foundation of China
  grantid: 62071445
GroupedDBID 0R~
6IK
97E
AAJGR
AASAJ
AAWTH
ABJNI
ABQJQ
ABVLG
AGQYO
AHBIQ
AKJIK
AKQYR
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
EBS
IFIPE
IPLJI
JAVBF
M43
OCL
PQQKQ
RIA
RIE
AAYXX
CITATION
7SC
8FD
ABAZT
JQ2
L7M
L~C
L~D
ID FETCH-LOGICAL-c293t-72818bfeeb19ea9795f8ca72110ef633000f800ebdab7972d70b62b6f124e3a33
IEDL.DBID RIE
ISICitedReferencesCount 9
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000979667300025&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2327-4697
IngestDate Mon Nov 10 02:48:14 EST 2025
Sat Nov 29 04:55:54 EST 2025
Tue Nov 18 22:05:36 EST 2025
Tue Nov 25 14:44:28 EST 2025
IsPeerReviewed true
IsScholarly true
Issue 3
Language English
License https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
https://doi.org/10.15223/policy-029
https://doi.org/10.15223/policy-037
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c293t-72818bfeeb19ea9795f8ca72110ef633000f800ebdab7972d70b62b6f124e3a33
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ORCID 0000-0001-8083-1805
0000-0003-2569-745X
0000-0002-1754-0607
0000-0002-3550-0625
PQID 2806216998
PQPubID 2040409
PageCount 11
ParticipantIDs crossref_citationtrail_10_1109_TNSE_2022_3227287
proquest_journals_2806216998
ieee_primary_9973351
crossref_primary_10_1109_TNSE_2022_3227287
PublicationCentury 2000
PublicationDate 2023-05-01
PublicationDateYYYYMMDD 2023-05-01
PublicationDate_xml – month: 05
  year: 2023
  text: 2023-05-01
  day: 01
PublicationDecade 2020
PublicationPlace Piscataway
PublicationPlace_xml – name: Piscataway
PublicationTitle IEEE transactions on network science and engineering
PublicationTitleAbbrev TNSE
PublicationYear 2023
Publisher IEEE
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml – name: IEEE
– name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References Konenỳ (ref4) 2016
ref13
Li (ref24) 2020
ref35
ref12
ref34
ref15
ref14
ref36
ref33
ref10
McMahan (ref2) 2017
Mnih (ref19) 2012
ref1
ref16
ref18
Ren (ref23) 2018
Wang (ref11) 2020; 33
Srivastava (ref17) 2014; 15
Arazo (ref30) 2019
ref26
ref25
Konenỳ (ref32) 2015
ref20
ref21
Goodfellow (ref9) 2016
Zhang (ref22) 2018
ref28
ref27
ref29
ref8
ref7
Yu (ref31) 2019
ref3
ref6
ref5
References_xml – ident: ref15
  doi: 10.1561/9781680837896
– volume: 15
  start-page: 1929
  issue: 1
  year: 2014
  ident: ref17
  article-title: Dropout: A simple way to prevent neural networks from overfitting
  publication-title: J. Mach. Learn. Res.
– ident: ref16
  doi: 10.1109/ICCV.2015.168
– ident: ref14
  doi: 10.1109/TWC.2020.2974748
– ident: ref7
  doi: 10.1109/TWC.2021.3099505
– ident: ref36
  doi: 10.1109/CVPR.2016.90
– ident: ref10
  doi: 10.1109/JIOT.2020.3023126
– start-page: 567
  volume-title: Proc. 29th Int. Conf. Mach. Learn.
  year: 2012
  ident: ref19
  article-title: Learning to label aerial images from noisy data
– ident: ref6
  doi: 10.1109/TNSE.2020.3014385
– start-page: 312
  volume-title: Proc. Int. Conf. Mach. Learn.
  year: 2019
  ident: ref30
  article-title: Unsupervised label noise modeling and loss correction
– ident: ref27
  doi: 10.1109/MWC.001.1900119
– ident: ref28
  doi: 10.1109/TNNLS.2022.3152527
– ident: ref18
  doi: 10.1109/CVPR.2015.7298885
– volume-title: Deep Learning
  year: 2016
  ident: ref9
– volume: 33
  start-page: 16070
  year: 2020
  ident: ref11
  article-title: Attack of the tails: Yes, you really can backdoor federated learning
  publication-title: Proc. Neural Inf. Process. Syst.
– ident: ref26
  doi: 10.1109/JIOT.2020.3027980
– ident: ref1
  doi: 10.1109/MCOM.2019.1900271
– ident: ref35
  doi: 10.1007/s10915-018-0757-z
– ident: ref25
  doi: 10.1109/TII.2019.2938861
– ident: ref3
  doi: 10.1145/3298981
– year: 2015
  ident: ref32
  article-title: Federated optimization: Distributed optimization beyond the datacenter
– ident: ref13
  doi: 10.1109/WCNC45663.2020.9120713
– ident: ref33
  doi: 10.1109/JSAC.2019.2904348
– ident: ref34
  doi: 10.1109/TAC.2014.2308612
– ident: ref21
  doi: 10.1609/aaai.v31i1.10894
– ident: ref8
  doi: 10.1109/TCOMM.2020.2979149
– ident: ref5
  doi: 10.1109/TNSE.2020.3016035
– start-page: 8778
  volume-title: Proc. Neural Inf. Process. Syst.
  year: 2018
  ident: ref22
  article-title: Generalized cross entropy loss for training deep neural networks with noisy labels
– start-page: 4334
  volume-title: Proc. Int. Conf. Mach. Learn.
  year: 2018
  ident: ref23
  article-title: Learning to reweight examples for robust deep learning
– year: 2016
  ident: ref4
  article-title: Federated learning: Strategies for improving communication efficiency
– volume-title: Proc. Int. Conf. Learn. Representations
  year: 2020
  ident: ref24
  article-title: Dividemix: Learning with noisy labels as semi-supervised learning
– ident: ref20
  doi: 10.1109/TGRS.2019.2961141
– ident: ref12
  doi: 10.1109/TCOMM.2021.3081746
– ident: ref29
  doi: 10.1561/9781680837896
– start-page: 7164
  volume-title: Proc. Int. Conf. Mach. Learn.
  year: 2019
  ident: ref31
  article-title: How does disagreement help generalization against label corruption?
– start-page: 1273
  volume-title: Proc. Artif. Intell. Statist.
  year: 2017
  ident: ref2
  article-title: Communication-efficient learning of deep networks from decentralized data
SSID ssj0001286333
Score 2.2916958
Snippet Federated learning (FL) is a communication-efficient machine learning paradigm to leverage distributed data at the network edge. Nevertheless, FL usually fails...
SourceID proquest
crossref
ieee
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 1
SubjectTerms Convergence
Data models
Datasets
Distributed networks
Federated learning
label noise
Loss measurement
Machine learning
Nodes
Noise measurement
non-convex optimization
parallel and distributed algorithms
robust design
Robustness
Servers
Training
Title Robust Federated Learning With Noisy Labeled Data Through Loss Function Correction
URI https://ieeexplore.ieee.org/document/9973351
https://www.proquest.com/docview/2806216998
Volume 10
WOSCitedRecordID wos000979667300025&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVIEE
  databaseName: IEEE Electronic Library (IEL)
  customDbUrl:
  eissn: 2334-329X
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0001286333
  issn: 2327-4697
  databaseCode: RIE
  dateStart: 20140101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3NS8MwFA9zeNCDX1OcTsnBk9itTdamOcq0eBhF5sTdStImOpBN1k7wv_clzcZAEbzlkEB4vyTvK-_3ELpSXIERzrlHhfK9fki0J304y0UQUBlrEtDAFgoPWZrGkwl_bKCbdS2MUsp-PlNdM7S5_GKeL02orMc5o9TUS28xxuparY14ShxRSl3iMvB5b5w-3YMDSEgXDi0j5tPchuqxvVR-PMBWqyT7_9vPAdpz1iO-reE-RA01O0K7G5yCLTQazeWyrHBiWCLAkCywo1B9xS_T6g2n82n5hYdCgr4p8J2oBB7XvXrwEHaNE1B0Biw8MH077PAYPSf348GD5xoneDlo78pjhuJJagXvMFeCMx7qOBfW11MapATi0mAoKlkIyTgjBfNlRGSkQdkrKig9Qc3ZfKZOETZ88ESGnPZZ3tdhJHUYExmEuQRPLGSijfyVTLPcsYqb5hbvmfUufJ4ZGDIDQ-ZgaKPr9ZKPmlLjr8ktI_f1RCfyNuqsgMvcpSszkyQmQQQO5Nnvq87RjukWX0dQOqhZLZbqAm3nn9W0XFza8_QNN0_H-A
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3fS8MwED7GFNQHf01xOjUPPond2qRtmkeZjom1iE70rSRtogPZZOsE_3uTtBsDRfAtDwmEuyR3l7v7PoAzyaR2whlzCJeu4wdYOcLVZzn3PCIihT3i2UbhmCZJ9PLC7mtwseiFkVLa4jPZNkOby8_H2cx8lXUYo4SYfumVwPexV3ZrLf2oRCEhpEpdei7rDJLHax0CYtzWx5ZiUza3ZHwsm8qPJ9jald7W_3a0DZuV_4guS4XvQE2OdmFjCVWwAQ8PYzGbFqhncCK0K5mjCkT1FT0PizeUjIfTLxRzoS1Ojq54wdGgZOtBsd416mlTZ9SFuoa5ww734Kl3Pej2nYo6wcm0_S4cakCehJL6JWaSM8oCFWXcRntSaSlpcSntKkqRc0EZxTl1RYhFqLS5l4QTsg_10XgkDwAZRHgsAkZ8mvkqCIUKIiy8IBM6Fgsob4I7l2maVbjiht7iPbXxhctSo4bUqCGt1NCE88WSjxJU46_JDSP3xcRK5E1ozRWXVtdumpo0MfZCHUIe_r7qFNb6g7s4jW-S2yNYN9zxZfViC-rFZCaPYTX7LIbTyYk9W9_ItctD
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Robust+Federated+Learning+With+Noisy+Labeled+Data+Through+Loss+Function+Correction&rft.jtitle=IEEE+transactions+on+network+science+and+engineering&rft.au=Chen%2C+Li&rft.au=Ang%2C+Fan&rft.au=Chen%2C+Yunfei&rft.au=Wang%2C+Weidong&rft.date=2023-05-01&rft.pub=The+Institute+of+Electrical+and+Electronics+Engineers%2C+Inc.+%28IEEE%29&rft.eissn=2334-329X&rft.volume=10&rft.issue=3&rft.spage=1501&rft_id=info:doi/10.1109%2FTNSE.2022.3227287&rft.externalDBID=NO_FULL_TEXT
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2327-4697&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2327-4697&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2327-4697&client=summon