1: \begin{abstract}

2: 

3: 

4: 

5: Classical federated learning (FL) enables training machine learning models without sharing data for privacy preservation, but heterogeneous data characteristic degrades the performance of the localized model. %

6: Personalized FL (PFL) addresses this by synthesizing personalized models from a global model via training on local data. %

7: Such a global model may overlook the specific information that the clients have been sampled. %

8: In this paper, we propose a novel scheme to inject personalized prior knowledge into the global model in each client, which attempts to mitigate the introduced incomplete information problem in PFL. %

9: At the heart of our proposed approach is a framework, the \textit{PFL with Bregman Divergence} (pFedBreD), decoupling the personalized prior from the local objective function regularized by Bregman divergence for greater adaptability in personalized scenarios.

10: We also relax the mirror descent (RMD) to extract the prior explicitly to provide optional strategies.

11: Additionally, our pFedBreD is backed up by a convergence analysis. %

12: Sufficient experiments demonstrate that our method reaches the \textit{state-of-the-art} performances on 5 datasets and outperforms other methods by up to 3.5\% across 8 benchmarks. 

13: Extensive analyses verify the robustness and necessity of proposed designs. %

14: \url{https://github.com/BDeMo/pFedBreD_public}

15: 

16: 

17: 

18: 

19: 

20: 

21: 

22: 

23: 

24: 

25: 

26: 

27: 

28: 

29: 

30: 

31: 

32: \end{abstract}

33: