The Panaceas for Improving Low-Rank Decomposition in Communication-Efficient Federated Learning

TL;DR

This paper introduces three novel low-rank decomposition techniques for federated learning to enhance training efficiency, convergence speed, and accuracy, supported by theoretical analysis and extensive experiments.

Contribution

It proposes three new decomposition methods—MUD, BKD, and AAD—that address key issues in FL and can be combined for improved performance.

Findings

Faster convergence compared to baseline methods

Superior accuracy demonstrated through experiments

Theoretical convergence guarantees for MUD

Abstract

To improve the training efficiency of federated learning (FL), previous research has employed low-rank decomposition techniques to reduce communication overhead. In this paper, we seek to enhance the performance of these low-rank decomposition methods. Specifically, we focus on three key issues related to decomposition in FL: what to decompose, how to decompose, and how to aggregate. Subsequently, we introduce three novel techniques: Model Update Decomposition (MUD), Block-wise Kronecker Decomposition (BKD), and Aggregation-Aware Decomposition (AAD), each targeting a specific issue. These techniques are complementary and can be applied simultaneously to achieve optimal performance. Additionally, we provide a rigorous theoretical analysis to ensure the convergence of the proposed MUD. Extensive experimental results show that our approach achieves faster convergence and superior accuracy compared to relevant baseline methods. The code is available at https://github.com/Leopold1423/fedmud-icml25.