Adaptive Gradient Sparsification for Efficient Federated Learning: An Online Learning Approach

Federated learning (FL) is an emerging technique for training machine learning models using geographically dispersed data collected by local entities. It includes local computation and synchronization steps. To reduce the communication overhead and improve the overall efficiency of FL, gradient spar...

Full description

Saved in:
Bibliographic Details
Published in:Proceedings of the International Conference on Distributed Computing Systems pp. 300 - 310
Main Authors: Han, Pengchao, Wang, Shiqiang, Leung, Kin K.
Format: Conference Proceeding
Language:English
Published: IEEE 01.11.2020
Subjects:
ISSN:2575-8411
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Federated learning (FL) is an emerging technique for training machine learning models using geographically dispersed data collected by local entities. It includes local computation and synchronization steps. To reduce the communication overhead and improve the overall efficiency of FL, gradient sparsification (GS) can be applied, where instead of the full gradient, only a small subset of important elements of the gradient is communicated. Existing work on GS uses a fixed degree of gradient sparsity for i.i.d.-distributed data within a datacenter. In this paper, we consider adaptive degree of sparsity and non-i.i.d. local datasets. We first present a fairness-aware GS method which ensures that different clients provide a similar amount of updates. Then, with the goal of minimizing the overall training time, we propose a novel online learning formulation and algorithm for automatically determining the near-optimal communication and computation trade-off that is controlled by the degree of gradient sparsity. The online learning algorithm uses an estimated sign of the derivative of the objective function, which gives a regret bound that is asymptotically equal to the case where exact derivative is available. Experiments with real datasets confirm the benefits of our proposed approaches, showing up to 40% improvement in model accuracy for a finite training time.
AbstractList Federated learning (FL) is an emerging technique for training machine learning models using geographically dispersed data collected by local entities. It includes local computation and synchronization steps. To reduce the communication overhead and improve the overall efficiency of FL, gradient sparsification (GS) can be applied, where instead of the full gradient, only a small subset of important elements of the gradient is communicated. Existing work on GS uses a fixed degree of gradient sparsity for i.i.d.-distributed data within a datacenter. In this paper, we consider adaptive degree of sparsity and non-i.i.d. local datasets. We first present a fairness-aware GS method which ensures that different clients provide a similar amount of updates. Then, with the goal of minimizing the overall training time, we propose a novel online learning formulation and algorithm for automatically determining the near-optimal communication and computation trade-off that is controlled by the degree of gradient sparsity. The online learning algorithm uses an estimated sign of the derivative of the objective function, which gives a regret bound that is asymptotically equal to the case where exact derivative is available. Experiments with real datasets confirm the benefits of our proposed approaches, showing up to 40% improvement in model accuracy for a finite training time.
Author Han, Pengchao
Leung, Kin K.
Wang, Shiqiang
Author_xml – sequence: 1
  givenname: Pengchao
  surname: Han
  fullname: Han, Pengchao
  email: hanpengchao199@gmail.com
  organization: Imperial College London,Department of Electrical and Electronic Engineering,UK
– sequence: 2
  givenname: Shiqiang
  surname: Wang
  fullname: Wang, Shiqiang
  email: wangshiq@us.ibm.com
  organization: IBM T. J. Watson Research Center,Yorktown Heights,NY,USA
– sequence: 3
  givenname: Kin K.
  surname: Leung
  fullname: Leung, Kin K.
  email: kin.leung@imperial.ac.uk
  organization: Imperial College London,Department of Electrical and Electronic Engineering,UK
BookMark eNo9jNtKw0AURUdRsKl-gQjzA6lzydx8C7GthUIfqq-W08wZHamTMAmCf9-i4tNm7bXZBblIXUJC7jibcc7c_ap5bLaVMaaaCSbYjDEm9BkpuBGWmxPYczIRyqjSVpxfkWIYPk4bZbWckNfaQz_GL6TLDD5iGum2hzzEEFsYY5do6DKdhxP-yAV6zDCip2uEnGJ6e6B1opt0iAn_O1r3fe6gfb8mlwEOA9785ZS8LObPzVO53ixXTb0uo2ByLKtq76UQ6FSQAcBpJ3XFvXPaokQrggQvFQRlvJZ279meW-ZYax20XgQtp-T29zci4q7P8RPy985JpYwz8giuYlYj
CODEN IEEPAD
ContentType Conference Proceeding
DBID 6IE
6IH
CBEJK
RIE
RIO
DOI 10.1109/ICDCS47774.2020.00026
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP) 1998-present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISBN 1728170028
9781728170022
EISSN 2575-8411
EndPage 310
ExternalDocumentID 9355797
Genre orig-research
GrantInformation_xml – fundername: Ministry of Defence
  funderid: 10.13039/100009941
GroupedDBID 23M
29G
29P
6IE
6IF
6IH
6IK
6IL
6IM
6IN
AAJGR
AAWTH
ABLEC
ACGFS
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IEGSK
IJVOP
IPLJI
M43
OCL
RIE
RIL
RIO
RNS
ID FETCH-LOGICAL-i203t-44bd322e95f3faa9693641d9968e3e82f3ad35af57d638bd0b18090c89acd2f63
IEDL.DBID RIE
ISICitedReferencesCount 127
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000667971400028&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Wed Aug 27 02:41:14 EDT 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i203t-44bd322e95f3faa9693641d9968e3e82f3ad35af57d638bd0b18090c89acd2f63
PageCount 11
ParticipantIDs ieee_primary_9355797
PublicationCentury 2000
PublicationDate 2020-Nov.
PublicationDateYYYYMMDD 2020-11-01
PublicationDate_xml – month: 11
  year: 2020
  text: 2020-Nov.
PublicationDecade 2020
PublicationTitle Proceedings of the International Conference on Distributed Computing Systems
PublicationTitleAbbrev ICDCS
PublicationYear 2020
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0005863
Score 2.561142
Snippet Federated learning (FL) is an emerging technique for training machine learning models using geographically dispersed data collected by local entities. It...
SourceID ieee
SourceType Publisher
StartPage 300
SubjectTerms Collaborative work
Distributed machine learning
edge computing
federated learning
Finite element analysis
gradient sparsification
Minimization
online learning
Optimization
Privacy
Synchronization
Training
Title Adaptive Gradient Sparsification for Efficient Federated Learning: An Online Learning Approach
URI https://ieeexplore.ieee.org/document/9355797
WOSCitedRecordID wos000667971400028&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
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV09T8MwED21FQNTgRbxLQ-MhCZ2EttsVWmBpapUkDpR-RN1SauQ8vuxndAysLBFTqJIZ539znnvHcCtQ8kS25RGKk5slCq3DnKciIhowam3zBKahWYTdDpliwWfteBup4UxxgTymbn3l-Ffvl6rrT8qG3gvcMppG9qU0lqrtadzsJw0Cp0k5oOX0eNonrrH_LkJ9vSt4J_wq4NK2EAm3f99-gj6eyUemu32mGNomeIEuj-tGFCTmT14H2qx8SsXeioDiatC842rWT0RKMQeOXCKxsEvwt-ceA8JBzM1agxWPx7QsEC18ehuDA0bx_E-vE3Gr6PnqGmdEK1wTKooTaV2qWp4ZokVguec5GmiXXHDDDEMWyI0yYTNqHYJKHUsvY9XrBgXSmObk1PoFOvCnAFybxPMHKzBrhLUggmVS5MlUjqgobJYnEPPh2u5qd0xlk2kLv4evoRDPx-1mu8KOlW5NddwoL6q1Wd5E6b0G265ouw
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PT8IwFH5BNNETKhh_24NHJ1vbba03giBEJCRgwknStR3hMggO_37bboIHL96WLsuS17z2e-33fQ_g3qDkBKc09qQfpB6VZh3kOBAeUYLH1jJLKOaaTcTDIZtO-agCD1stjNbakc_0o310d_lqKTf2qKxpvcBjHu_BfkgpDgq11o7QwSJSanQCnzf77ef2mMYG3pgqEFsCl3NQ-NVDxW0h3dr_fn4MjZ0WD422u8wJVHR2CrWfZgyozM06fLSUWNm1C72sHY0rR-OVqVotFchFHxl4ijrOMcK-7FoXCQM0FSotVudPqJWhwnp0O4Zaped4A967nUm755XNE7wF9knuUZook6yahylJheARJxENlClvmCaa4ZQIRUKRhrEyKZgoP7FOXr5kXEiF04icQTVbZvockPmaYGaADTa1oBJMyCjRYZAkBmrI0BcXULfhmq0Kf4xZGanLv4fv4LA3eRvMBv3h6xUc2bkptH3XUM3XG30DB_IrX3yub930fgMVyaYz
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=proceeding&rft.title=Proceedings+of+the+International+Conference+on+Distributed+Computing+Systems&rft.atitle=Adaptive+Gradient+Sparsification+for+Efficient+Federated+Learning%3A+An+Online+Learning+Approach&rft.au=Han%2C+Pengchao&rft.au=Wang%2C+Shiqiang&rft.au=Leung%2C+Kin+K.&rft.date=2020-11-01&rft.pub=IEEE&rft.eissn=2575-8411&rft.spage=300&rft.epage=310&rft_id=info:doi/10.1109%2FICDCS47774.2020.00026&rft.externalDocID=9355797