A Huber Loss Minimization Approach to Byzantine Robust Federated Learning
This addresses security vulnerabilities in federated learning systems, which is critical for privacy-preserving distributed machine learning, but appears incremental as it builds on existing robust aggregation methods.
The paper tackles the problem of adversarial attacks in federated learning by introducing a novel aggregator based on Huber loss minimization, achieving optimal dependence on the attack ratio and not requiring precise knowledge of it, with advantages under both i.i.d. and non-i.i.d. data conditions.
Federated learning systems are susceptible to adversarial attacks. To combat this, we introduce a novel aggregator based on Huber loss minimization, and provide a comprehensive theoretical analysis. Under independent and identically distributed (i.i.d) assumption, our approach has several advantages compared to existing methods. Firstly, it has optimal dependence on $ε$, which stands for the ratio of attacked clients. Secondly, our approach does not need precise knowledge of $ε$. Thirdly, it allows different clients to have unequal data sizes. We then broaden our analysis to include non-i.i.d data, such that clients have slightly different distributions.