Skip to main navigation Skip to search Skip to main content

Federated model search via reinforcement learning

  • Dixi Yao
  • , Lingdong Wang
  • , Jiayu Xu
  • , Liyao Xiang*
  • , Shuo Shao
  • , Yingqi Chen
  • , Yanjun Tong
  • *Corresponding author for this work
  • Shanghai Jiao Tong University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Federated Learning (FL) framework enables training over distributed datasets while keeping the data local. However, it is difficult to customize a model fitting for all unknown local data. A pre-determined model is most likely to lead to slow convergence or low accuracy, especially when the distributed data is non-i.i.d.. To resolve the issue, we propose a model searching method in the federated learning scenario, and the method automatically searches a model structure fitting for the unseen local data. We novelly design a reinforcement learning-based framework that samples and distributes sub-models to the participants and updates its model selection policy by maximizing the reward. In practice, the model search algorithm takes a long time to converge, and hence we adaptively assign sub-models to participants according to the transmission condition. We further propose delay-compensated synchronization to mitigate loss over late updates to facilitate convergence. Extensive experiments show that our federated model search algorithm produces highly accurate models efficiently, particularly on non-i.i.d. data.

Original languageEnglish
Title of host publicationProceedings - 2021 IEEE 41st International Conference on Distributed Computing Systems, ICDCS 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages830-840
Number of pages11
ISBN (Electronic)9781665445139
DOIs
StatePublished - Jul 2021
Externally publishedYes
Event41st IEEE International Conference on Distributed Computing Systems, ICDCS 2021 - Virtual, Online, United States
Duration: 7 Jul 2021 → 10 Jul 2021

Publication series

NameProceedings - International Conference on Distributed Computing Systems
Volume2021-July
ISSN (Print)1063-6927
ISSN (Electronic)2575-8411

Conference

Conference41st IEEE International Conference on Distributed Computing Systems, ICDCS 2021
Country/TerritoryUnited States
CityVirtual, Online
Period7/07/21 → 10/07/21

Keywords

  • Edge Cloud Computing
  • Neural Architecture Search
  • Reinforcement Learning

Fingerprint

Dive into the research topics of 'Federated model search via reinforcement learning'. Together they form a unique fingerprint.

Cite this