id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
fcis-16180	Zhou, Qun	PPBRFL: Privacy-Preserving Byzantine-Robust Federated Learning	2024	7	.pdf	application/pdf	5276	345	59	Furthermore, we introduce a reward and penalty mechanism that considers users’ behavior to mitigate the impact of Byzantine users on the global model. Byzantine users can upload false parameters to the cloud server, thereby compromising global model’s accuracy or manipulating the global model to produce specific outputs through poisoning attacks [10].	cache/fcis-16180.pdf	txt/fcis-16180.txt