BriefGPT.xyz
Jul, 2022
FedDM:基于迭代分布匹配的通信高效联邦学习
FedDM: Iterative Distribution Matching for Communication-Efficient Federated Learning
HTML
PDF
Yuanhao Xiong, Ruochen Wang, Minhao Cheng, Felix Yu, Cho-Jui Hsieh
TL;DR
本研究提出了FedDM,旨在通过多个本地替代函数来构建全局训练目标,从而减少通信轮数,改善模型质量,并在保留差分隐私的同时证明该算法的有效性。
Abstract
federated learning
~(
fl
) has recently attracted increasing attention from academia and industry, with the ultimate goal of achieving collaborative training under privacy and
→