EDEN: Communication-Efficient and Robust Distributed Mean Estimation for Federated Learning

TL;DR

EDEN is a novel distributed mean estimation method designed for federated learning that effectively manages communication constraints and packet losses, improving accuracy and robustness over existing techniques.

Contribution

We introduce EDEN, a robust DME technique that handles heterogeneous communication budgets and packet losses with proven theoretical guarantees.

Findings

EDEN outperforms state-of-the-art DME methods in various settings.

EDEN provides strong theoretical guarantees for estimation accuracy.

Empirical results show improved learning performance with EDEN.

Abstract

Distributed Mean Estimation (DME) is a central building block in federated learning, where clients send local gradients to a parameter server for averaging and updating the model. Due to communication constraints, clients often use lossy compression techniques to compress the gradients, resulting in estimation inaccuracies. DME is more challenging when clients have diverse network conditions, such as constrained communication budgets and packet losses. In such settings, DME techniques often incur a significant increase in the estimation error leading to degraded learning performance. In this work, we propose a robust DME technique named EDEN that naturally handles heterogeneous communication budgets and packet losses. We derive appealing theoretical guarantees for EDEN and evaluate it empirically. Our results demonstrate that EDEN consistently improves over state-of-the-art DME techniques.