DPFL: Decentralized Personalized Federated Learning
Kharrat, Salma Canini, Marco Horváth, Samuel
Kharrat, Salma
Canini, Marco
Horváth, Samuel
Supervisor
Department
Machine Learning
Embargo End Date
Type
Conference proceeding
Date
2025
License
Language
English
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
This work addresses the challenges of data heterogeneity and communication constraints in decentralized federated learning (FL). We introduce decentralized personalized FL (DPFL), a bi-level optimization framework that enhances personalized FL by leveraging combinatorial relationships among clients, enabling fine-grained and targeted collaborations. By employing a constrained greedy algorithm, DPFL constructs a collaboration graph that guides clients in choosing suitable collaborators, enabling personalized model training tailored to local data while respecting a fixed and predefined communication and resource budget. Our theoretical analysis demonstrates that the proposed objective for constructing the collaboration graph yields superior or equivalent performance compared to any alternative collaboration structures, including pure local training. Extensive experiments across diverse datasets show that DPFL consistently outperforms existing methods, effectively handling non-IID data, reducing communication overhead, and improving resource efficiency in real-world decentralized FL scenarios. The code can be accessed at: https://github.com/salmakh1/DPFL.
Citation
S. Kharrat, M. Canini, and S. Horváth, “DPFL: Decentralized Personalized Federated Learning,” in Proc. 28th Int. Conf. Artif. Intell. Stat. (AISTATS 2025), vol. 258, Proc. Mach. Learn. Res., Mai Khao, Thailand, May 3–5, 2025, pp. 5086–5094.
Source
Proceedings of Machine Learning Research
Conference
28th International Conference on Artificial Intelligence and Statistics, AISTATS 2025
Keywords
Data Accuracy, Data Handling, Decentralized Finance, Decentralized Systems, Federated Learning, Graph Algorithms, Optimization, Bi-level Optimization, Communication Constraints, Data Heterogeneity, Data-communication, Decentralised, Fine Grained, Greedy Algorithms, Model Training, Optimization Framework, Personalized Model, Budget Control
Subjects
Source
28th International Conference on Artificial Intelligence and Statistics, AISTATS 2025
Publisher
ML Research Press