1. we find that this line is close to NP-FedAvg and has higher model accuracy than the FedRtid, but the privacy guarantee of DP-FedAvg2 is much less than that of FedRtid;DP
2. Communication-efficient learning of deep networks from decentralized data;B Mcmahan;Proceedings of the 20th International Conference on Artificial Intelligence and Statistics,2017
3. Figure 3: Under IID and Non-IID data distribution, the total time taken to complete the training of the model and the test accuracy rate;Z Li;Nature Machine Intelligence,2021