Enhancing Robustness of Federated Learning via Server Learning

TL;DR

This paper proposes a server learning approach with client update filtering and geometric median aggregation to improve federated learning robustness against malicious clients, especially under non-i.i.d. data distributions.

Contribution

It introduces a heuristic algorithm combining server learning and filtering to significantly enhance model accuracy in adversarial federated learning scenarios.

Findings

Achieves high accuracy even with over 50% malicious clients

Effective with small, synthetic server datasets

Improves robustness against non-i.i.d. data distributions

Abstract

This paper explores the use of server learning for enhancing the robustness of federated learning against malicious attacks even when clients' training data are not independent and identically distributed. We propose a heuristic algorithm that uses server learning and client update filtering in combination with geometric median aggregation. We demonstrate via experiments that this approach can achieve significant improvement in model accuracy even when the fraction of malicious clients is high, even more than $50\%$ in some cases, and the dataset utilized by the server is small and could be synthetic with its distribution not necessarily close to that of the clients' aggregated data.