Secure and accurate personalized federated learning with similarity-based model aggregation

Personalized federated learning (PFL) combines client needs and data characteristics to train personalized models for local clients. However, the most of previous PFL schemes encountered challenges such as low model prediction accuracy and privacy leakage when applied to practical datasets. Besides,...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on sustainable computing 2024, p.1-14
Main Authors: Tan, Zhouyong, Le, Junqing, Yang, Fan, Huang, Min, Xiang, Tao, Liao, Xiaofeng
Format: Article
Language:English
Subjects:
Adaptation models
Computational modeling
Data models
Federated learning
Personalized federated learning
Predictive models
Privacy
Privacy protection
Secure aggregation
Servers
Similarity metric
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Personalized federated learning (PFL) combines client needs and data characteristics to train personalized models for local clients. However, the most of previous PFL schemes encountered challenges such as low model prediction accuracy and privacy leakage when applied to practical datasets. Besides, the existing privacy protection methods fail to achieve satisfactory results in terms of model prediction accuracy and security simultaneously. In this paper, we propose a Privacy-preserving Personalized Federated Learning under Secure Multi-party Computation (SMC-PPFL), which can preserve privacy while obtaining a local personalized model with high prediction accuracy. In SMC-PPFL, noise perturbation is utilized to protect similarity computation, and secure multi-party computation is employed for model sub-aggregations. This combination ensures that clients' privacy is preserved, and the computed values remain unbiased without compromising security. Then, we propose a weighted sub-aggregation strategy based on the similarity of clients and introduce a regularization term in the local training to improve prediction accuracy. Finally, we evaluate the performance of SMC-PPFL on three common datasets. The experimental results show that SMC-PPFL achieves 2% ∼ 15% higher prediction accuracy compared to the previous PFL schemes. Besides, the security analysis also verifies that SMC-PPFL can resist model inversion attacks and membership inference attacks
ISSN:2377-3782
2377-3782
2377-3790
DOI:10.1109/TSUSC.2024.3403427