How Does the Smoothness Approximation Method Facilitate Generalization for Federated Adversarial Learning?

Open in new window