FedImpro: Measuring and Improving Client Update in Federated Learning

TL;DR

FedImpro introduces a novel method for mitigating client drift in federated learning by reconstructing feature distributions and decoupling model components, leading to improved generalization across heterogeneous data sources.

Contribution

This paper presents FedImpro, a new approach that constructs similar conditional distributions to reduce client dissimilarity and improve federated learning performance.

Findings

FedImpro reduces gradient dissimilarity in FL.

It enhances model generalization under data heterogeneity.

Experimental results confirm improved performance with FedImpro.

Abstract

Federated Learning (FL) models often experience client drift caused by heterogeneous data, where the distribution of data differs across clients. To address this issue, advanced research primarily focuses on manipulating the existing gradients to achieve more consistent client models. In this paper, we present an alternative perspective on client drift and aim to mitigate it by generating improved local models. First, we analyze the generalization contribution of local training and conclude that this generalization contribution is bounded by the conditional Wasserstein distance between the data distribution of different clients. Then, we propose FedImpro, to construct similar conditional distributions for local training. Specifically, FedImpro decouples the model into high-level and low-level components, and trains the high-level portion on reconstructed feature distributions. This approach enhances the generalization contribution and reduces the dissimilarity of gradients in FL. Experimental results show that FedImpro can help FL defend against data heterogeneity and enhance the generalization performance of the model.