Asynchronous Federated Optimization

Cong Xie; Sanmi Koyejo; Indranil Gupta

Asynchronous Federated Optimization

Distributed, Parallel, and Cluster Computing 2020-12-08 v5 Machine Learning

Authors: Cong Xie , Sanmi Koyejo , Indranil Gupta

Abstract

Federated learning enables training on a massive number of edge devices. To improve flexibility and scalability, we propose a new asynchronous federated optimization algorithm. We prove that the proposed approach has near-linear convergence to a global optimum, for both strongly convex and a restricted family of non-convex problems. Empirical results show that the proposed algorithm converges quickly and tolerates staleness in various applications.

Keywords

convex optimization federated learning optimization

Cite

@article{arxiv.1903.03934,
  title  = {Asynchronous Federated Optimization},
  author = {Cong Xie and Sanmi Koyejo and Indranil Gupta},
  journal= {arXiv preprint arXiv:1903.03934},
  year   = {2020}
}

Related papers

View all related →