Skip to main navigation Skip to search Skip to main content

EdgePS: Selective Parameter Aggregation for Distributed Machine Learning in Edge Computing

  • University of Science and Technology of China
  • California State University San Bernardino

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

In this paper, we propose EdgePS, an advanced parameter server approach for distributed machine learning in edge computing scenarios. Different from the Conventional Parameter Server (CPS) approach, which performs parameter aggregation after every local training epoch, EdgePS synchronizes the parameters of all workers only when the local training cannot improve the global model performance. We first analyze how the local training will impact the performance of the global model, and then design algorithms to determine when the best time is to perform the parameter aggregation. Both real testbed experiments and extensive large scale simulations demonstrate that EdgePS can train a practical machine learning model, e.g., VGG-16, with up to 59.28% less time compared with the CPS approach. With the same training time, EdgePS can improve model accuracy by up to 30.19 % compared with the state-of-The-Art distributed machine learning algorithm designed for edge computing scenarios.

Original languageEnglish
Title of host publicationProceedings - 2021 IEEE 14th International Conference on Cloud Computing, CLOUD 2021
EditorsClaudio Agostino Ardagna, Carl K. Chang, Ernesto Daminai, Rajiv Ranjan, Zhongjie Wang, Robert Ward, Jia Zhang, Wensheng Zhang
PublisherIEEE Computer Society
Pages217-227
Number of pages11
ISBN (Electronic)9781665400602
DOIs
StatePublished - Sep 2021
Event14th IEEE International Conference on Cloud Computing, CLOUD 2021 - Virtual, Online, United States
Duration: Sep 5 2021Sep 11 2021

Publication series

NameIEEE International Conference on Cloud Computing, CLOUD
Volume2021-September
ISSN (Print)2159-6182
ISSN (Electronic)2159-6190

Conference

Conference14th IEEE International Conference on Cloud Computing, CLOUD 2021
Country/TerritoryUnited States
CityVirtual, Online
Period09/5/2109/11/21

Keywords

  • n/a

Fingerprint

Dive into the research topics of 'EdgePS: Selective Parameter Aggregation for Distributed Machine Learning in Edge Computing'. Together they form a unique fingerprint.

Cite this