FedCBO: Reaching Group Consensus in Clustered Federated Learning through Consensus-based Optimization

TL;DR

FedCBO introduces a consensus-based optimization approach for clustered federated learning, enabling effective group-specific model training without prior group knowledge, supported by theoretical analysis and empirical validation.

Contribution

The paper presents a novel CBO-inspired method for clustered federated learning that does not require group membership information and provides convergence guarantees.

Findings

FedCBO outperforms existing methods in experiments.

Theoretical analysis confirms convergence in non-convex settings.

Method effectively identifies group-specific models without prior group labels.

Abstract

Federated learning is an important framework in modern machine learning that seeks to integrate the training of learning models from multiple users, each user having their own local data set, in a way that is sensitive to data privacy and to communication loss constraints. In clustered federated learning, one assumes an additional unknown group structure among users, and the goal is to train models that are useful for each group, rather than simply training a single global model for all users. In this paper, we propose a novel solution to the problem of clustered federated learning that is inspired by ideas in consensus-based optimization (CBO). Our new CBO-type method is based on a system of interacting particles that is oblivious to group memberships. Our model is motivated by rigorous mathematical reasoning, including a mean field analysis describing the large number of particles limit of our particle system, as well as convergence guarantees for the simultaneous global optimization of general non-convex objective functions (corresponding to the loss functions of each cluster of users) in the mean-field regime. Experimental results demonstrate the efficacy of our FedCBO algorithm compared to other state-of-the-art methods and help validate our methodological and theoretical work.