1 paper · 1 filter
Tianyun Zhang, Zhen Yang, Haozhao Wang +2
Federated learning faces increasing threats from model poisoning attacks, which harms its application to improve privacy. Existing defense methods typically rely on fixed threshold…