English

LoSAC: An Efficient Local Stochastic Average Control Method for Federated Optimization

Distributed, Parallel, and Cluster Computing 2023-02-28 v3 Machine Learning

Abstract

Federated optimization (FedOpt), which targets at collaboratively training a learning model across a large number of distributed clients, is vital for federated learning. The primary concerns in FedOpt can be attributed to the model divergence and communication efficiency, which significantly affect the performance. In this paper, we propose a new method, i.e., LoSAC, to learn from heterogeneous distributed data more efficiently. Its key algorithmic insight is to locally update the estimate for the global full gradient after {each} regular local model update. Thus, LoSAC can keep clients' information refreshed in a more compact way. In particular, we have studied the convergence result for LoSAC. Besides, the bonus of LoSAC is the ability to defend the information leakage from the recent technique Deep Leakage Gradients (DLG). Finally, experiments have verified the superiority of LoSAC comparing with state-of-the-art FedOpt algorithms. Specifically, LoSAC significantly improves communication efficiency by more than 100%100\% on average, mitigates the model divergence problem and equips with the defense ability against DLG.

Keywords

Cite

@article{arxiv.2112.07839,
  title  = {LoSAC: An Efficient Local Stochastic Average Control Method for Federated Optimization},
  author = {Huiming Chen and Huandong Wang and Quanming Yao and Yong Li and Depeng Jin and Qiang Yang},
  journal= {arXiv preprint arXiv:2112.07839},
  year   = {2023}
}

Comments

ACM Transactions on Knowledge Discovery from Datahttps://doi.org/10.1145/3566128

R2 v1 2026-06-24T08:17:44.664Z