Communication-Efficient ADMM-based Federated Learning
arXiv:2110.15318
Abstract
Federated learning has shown its advances over the last few years but is facing many challenges, such as how algorithms save communication resources, how they reduce computational costs, and whether they converge. To address these issues, this paper proposes exact and inexact ADMM-based federated learning. They are not only communication-efficient but also converge linearly under very mild conditions, such as convexity-free and irrelevance to data distributions. Moreover, the inexact version has low computational complexity, thereby alleviating the computational burdens significantly.
References in corpus (9)
- Federated Optimization: Distributed Machine Learning for On-Device Intelligence
- Federated Optimization:Distributed Optimization Beyond the Datacenter
- Distributed Federated Learning for Ultra-Reliable Low-Latency Vehicular Communications
- Asynchronous Stochastic Gradient Descent with Delay Compensation
- Improving the Privacy and Accuracy of ADMM-Based Distributed Algorithms
- Inexact-ADMM Based Federated Meta-Learning for Fast and Continual Edge Learning
- Communication Efficient Distributed Learning with Censored, Quantized, and Generalized Group ADMM
- Federated Nonconvex Sparse Learning
- Differentially Private Federated Learning via Inexact ADMM