Communication-Efficient Federated AUC Maximization with Cyclic Client Participation

TL;DR

This paper develops communication-efficient algorithms for federated AUC maximization with cyclic client participation, addressing practical constraints and providing theoretical guarantees and empirical validation.

Contribution

It introduces novel algorithms with provable communication and iteration complexity bounds for federated AUC maximization under cyclic client participation, including the PL condition.

Findings

Achieves state-of-the-art communication complexity of  O(1/^{1/2}) for squared surrogate loss.

Provides theoretical complexity bounds for general pairwise AUC losses.

Demonstrates superior empirical performance on benchmark tasks.

Abstract

Federated AUC maximization is a powerful approach for learning from imbalanced data in federated learning (FL). However, existing methods typically assume full client availability, which is rarely practical. In real-world FL systems, clients often participate in a cyclic manner: joining training according to a fixed, repeating schedule. This setting poses unique optimization challenges for the non-decomposable AUC objective. This paper addresses these challenges by developing and analyzing communication-efficient algorithms for federated AUC maximization under cyclic client participation. We investigate two key settings: First, we study AUC maximization with a squared surrogate loss, which reformulates the problem as a nonconvex-strongly-concave minimax optimization. By leveraging the Polyak-{\L}ojasiewicz (PL) condition, we establish a state-of-the-art communication complexity of $\widetilde{O}(1/\epsilon^{1/2})$ and iteration complexity of $\widetilde{O}(1/\epsilon)$. Second, we consider general pairwise AUC losses. We establish a communication complexity of $O(1/\epsilon^3)$ and an iteration complexity of $O(1/\epsilon^4)$. Further, under the PL condition, these bounds improve to communication complexity of $\widetilde{O}(1/\epsilon^{1/2})$ and iteration complexity of $\widetilde{O}(1/\epsilon)$. Extensive experiments on benchmark tasks in image classification, medical imaging, and fraud detection demonstrate the superior efficiency and effectiveness of our proposed methods.