Paper Overview
Research Area: ML Authors: Van Sy Mai, Kushal Chakrabarti, Richard J. La, et al. Published: 2026-04-03 arXiv: 2604.03226
Summary
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 (non-IID).
Approach
The authors propose a heuristic algorithm that uses:
- Server learning — leveraging a small auxiliary dataset at the server
- Client update filtering — removing suspicious updates before aggregation
- Geometric median aggregation — a robust aggregation rule
- The approach achieves significant improvement in model accuracy even when the fraction of malicious clients is high — exceeding 50% in some cases.
- It remains effective even when the server's dataset is small and could be synthetic, with a distribution not necessarily close to that of the clients' aggregated data.
Key Findings
Original 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.
--- *Auto-collected on 2026-04-06*