Federated Learning for Distribution Skewed Data Using Sample Weights

One of the most challenging issues in federated learning (FL) is that the data are often not independent and identically distributed (non-IID). Clients are expected to contribute the same type of data and drawn from one global distribution. However, data are often collected in different ways from different resources. Thus, the data distributions among clients might be different from the underlying global distribution. This creates a weight divergence issue and reduces FL performance. This work focuses on improving FL performance for skewed data distribution across clients. The main idea is to adjust the client distribution closer to the global distribution using sample weights. Thus, the machine learning (ML) model converges faster with higher accuracy. We start from the fundamental concept of empirical risk minimization and theoretically derive a solution for adjusting the distribution skewness using sample weights. To determine sample weights, we implicitly exchange density information by leveraging a neural network-based density estimation model, MADE. The clients’ data distribution can then be adjusted without exposing their raw data. Our experiment results on three real-world datasets show that the proposed method not only improves FL accuracy but also significantly reduces communication costs compared to the other experimental methods.