FedSoft: Soft Clustered Federated Learning with Proximal Local Updating

TL;DR

FedSoft introduces a soft clustering approach in federated learning, enabling clients with mixed data distributions to collaboratively learn personalized and cluster models efficiently, with limited client workload per round.

Contribution

It proposes FedSoft, a novel method that relaxes hard clustering assumptions, allowing for mixed data distributions and efficient proximal updates in federated learning.

Findings

FedSoft effectively exploits distribution similarities.

It learns personalized and cluster models with high accuracy.

Requires only one optimization task per client per round.

Abstract

Traditionally, clustered federated learning groups clients with the same data distribution into a cluster, so that every client is uniquely associated with one data distribution and helps train a model for this distribution. We relax this hard association assumption to soft clustered federated learning, which allows every local dataset to follow a mixture of multiple source distributions. We propose FedSoft, which trains both locally personalized models and high-quality cluster models in this setting. FedSoft limits client workload by using proximal updates to require the completion of only one optimization task from a subset of clients in every communication round. We show, analytically and empirically, that FedSoft effectively exploits similarities between the source distributions to learn personalized and cluster models that perform well.