May, 2024
使用更新摘要和基于投票的防御提升联邦学习的安全和隐私性
Enhancing Security and Privacy in Federated Learning using Update Digests and Voting-Based Defense
Wenjie Li, Kai Fan, Jingyuan Zhang, Hui Li, Wei Yang Bryan Lim...
TL;DRFederated Learning with Update Digest (FLUD) is a novel framework that ensures privacy preservation, resistance to Byzantine attacks, and reduced overhead through the use of the LinfSample method and optimized SMPC protocols.