Asynchronous Federated Learning Through Online Linear Regressions

In the practical scenario of Federated Learning (FL), clients upload their local model to a server at different times owing to heterogeneity in the clients’ device environment. Therefore, Asynchronous Federated Learning (AFL) has been aggressively studied recently. Although the initial mo...

Full description

Saved in:
Bibliographic Details
Main Authors: Taiga Kashima, Ayako Amma, Hideki Nakayama
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10811896/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846102112433340416
author Taiga Kashima
Ayako Amma
Hideki Nakayama
author_facet Taiga Kashima
Ayako Amma
Hideki Nakayama
author_sort Taiga Kashima
collection DOAJ
description In the practical scenario of Federated Learning (FL), clients upload their local model to a server at different times owing to heterogeneity in the clients’ device environment. Therefore, Asynchronous Federated Learning (AFL) has been aggressively studied recently. Although the initial motivation for AFL is to reduce the difficulties of FL, AFL itself also has problems in practice (that is, the latest uploaded local model affects the performance of the global model). As a scientific challenge, we would like to develop a simple yet broadly applicable concept for AFL; thus, we revisited a classic machine learning theory and found a high conceptual affinity between AFL and online linear regression—both are problem formulations for sequential inputs. On the basis of this underlying philosophy, we propose a framework for AFL in which classic online linear regression techniques are utilized in server aggregation to realize an asynchronous global model update. Our framework can provide a convergence guarantee of the global model performance relative to a synchronized aggregation method through the theoretical guarantee in online linear regressions. The experiments show the superiority of our asynchronous updates in a server in both image and text tasks under non-IID data.
format Article
id doaj-art-1b6f4a109c66479eadda50e27dfd24d4
institution Kabale University
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-1b6f4a109c66479eadda50e27dfd24d42024-12-28T00:00:52ZengIEEEIEEE Access2169-35362024-01-011219513119514410.1109/ACCESS.2024.352100910811896Asynchronous Federated Learning Through Online Linear RegressionsTaiga Kashima0https://orcid.org/0000-0002-5709-4752Ayako Amma1Hideki Nakayama2Graduate School of Information Science and Technology, The University of Tokyo, Tokyo, JapanWoven by Toyota Inc., Tokyo, JapanGraduate School of Information Science and Technology, The University of Tokyo, Tokyo, JapanIn the practical scenario of Federated Learning (FL), clients upload their local model to a server at different times owing to heterogeneity in the clients’ device environment. Therefore, Asynchronous Federated Learning (AFL) has been aggressively studied recently. Although the initial motivation for AFL is to reduce the difficulties of FL, AFL itself also has problems in practice (that is, the latest uploaded local model affects the performance of the global model). As a scientific challenge, we would like to develop a simple yet broadly applicable concept for AFL; thus, we revisited a classic machine learning theory and found a high conceptual affinity between AFL and online linear regression—both are problem formulations for sequential inputs. On the basis of this underlying philosophy, we propose a framework for AFL in which classic online linear regression techniques are utilized in server aggregation to realize an asynchronous global model update. Our framework can provide a convergence guarantee of the global model performance relative to a synchronized aggregation method through the theoretical guarantee in online linear regressions. The experiments show the superiority of our asynchronous updates in a server in both image and text tasks under non-IID data.https://ieeexplore.ieee.org/document/10811896/Asynchronous federated learningonline linear regressionfederated learning
spellingShingle Taiga Kashima
Ayako Amma
Hideki Nakayama
Asynchronous Federated Learning Through Online Linear Regressions
IEEE Access
Asynchronous federated learning
online linear regression
federated learning
title Asynchronous Federated Learning Through Online Linear Regressions
title_full Asynchronous Federated Learning Through Online Linear Regressions
title_fullStr Asynchronous Federated Learning Through Online Linear Regressions
title_full_unstemmed Asynchronous Federated Learning Through Online Linear Regressions
title_short Asynchronous Federated Learning Through Online Linear Regressions
title_sort asynchronous federated learning through online linear regressions
topic Asynchronous federated learning
online linear regression
federated learning
url https://ieeexplore.ieee.org/document/10811896/
work_keys_str_mv AT taigakashima asynchronousfederatedlearningthroughonlinelinearregressions
AT ayakoamma asynchronousfederatedlearningthroughonlinelinearregressions
AT hidekinakayama asynchronousfederatedlearningthroughonlinelinearregressions