Sparse Training for Federated Learning with Regularized Error Correction

Ran Greidi, Kobi Cohen

Federated Learning (FL) has attracted much interest due to the significant advantages it brings to training deep neural network (DNN) models. However, since communications and comp…