Federated Communication-Efficient Multi-Objective Optimization

TL;DR

This paper introduces FedCMOO, a communication-efficient federated multi-objective optimization algorithm that converges faster and scales better with objectives, allowing user preferences and demonstrating superior performance.

Contribution

The paper proposes FedCMOO, a novel federated multi-objective optimization method that reduces communication costs and provides convergence guarantees under milder assumptions.

Findings

FedCMOO outperforms baseline methods in convergence speed.

Communication cost of FedCMOO does not increase with the number of objectives.

The method allows user-specified preferences over objectives.

Abstract

We study a federated version of multi-objective optimization (MOO), where a single model is trained to optimize multiple objective functions. MOO has been extensively studied in the centralized setting but is less explored in federated or distributed settings. We propose FedCMOO, a novel communication-efficient federated multi-objective optimization (FMOO) algorithm that improves the error convergence performance of the model compared to existing approaches. Unlike prior works, the communication cost of FedCMOO does not scale with the number of objectives, as each client sends a single aggregated gradient to the central server. We provide a convergence analysis of the proposed method for smooth and non-convex objective functions under milder assumptions than in prior work. In addition, we introduce a variant of FedCMOO that allows users to specify a preference over the objectives in terms of a desired ratio of the final objective values. Through extensive experiments, we demonstrate the superiority of our proposed method over baseline approaches.