Rethinking Data Heterogeneity in Federated Learning: Introducing a New Notion and Standard Benchmarks

Open in new window