5 papers
Divergence-Based Adaptive Aggregation for Byzantine Robust Federated Learning
Bingnan Xiao, Feng Zhu, Jingjing Zhang +2
Inherent client drifts caused by data heterogeneity, as well as vulnerability to Byzantine attacks within the system, hinder effective model training and convergence in federated l…
DRAG: Divergence-based Adaptive Aggregation in Federated learning on Non-IID Data
Feng Zhu, Jingjing Zhang, Shengyun Liu +1
Local stochastic gradient descent (SGD) is a fundamental approach in achieving communication efficiency in Federated Learning (FL) by allowing individual workers to perform local u…
Communication-Efficient Local SGD with Age-Based Worker Selection
Feng Zhu, Jingjing Zhang, Xin Wang
A major bottleneck of distributed learning under parameter-server (PS) framework is communication cost due to frequent bidirectional transmissions between the PS and workers. To ad…
STSyn: Speeding Up Local SGD with Straggler-Tolerant Synchronization
Feng Zhu, Jingjing Zhang, Xin Wang
Synchronous local stochastic gradient descent (local SGD) suffers from some workers being idle and random delays due to slow and straggling workers, as it waits for the workers to…
Adaptive Worker Grouping For Communication-Efficient and Straggler-Tolerant Distributed SGD
Feng Zhu, Jingjing Zhang, Osvaldo Simeone +1
Wall-clock convergence time and communication load are key performance metrics for the distributed implementation of stochastic gradient descent (SGD) in parameter server settings.…