Optimal Client Sampling in Federated Learning with Client-Level Heterogeneous Differential Privacy

Open in new window