Distributionally Robust Federated Averaging

Neural Information Processing Systems 

In this paper, we study communication efficient distributed algorithms for distribu-tionally robust federated learning via periodic averaging with adaptive sampling.