The Science and Information (SAI) Organization
  • Home
  • About Us
  • Journals
  • Conferences
  • Contact Us

Publication Links

  • IJACSA
  • Author Guidelines
  • Publication Policies
  • Digital Archiving Policy
  • Promote your Publication
  • Metadata Harvesting (OAI2)

IJACSA

  • About the Journal
  • Call for Papers
  • Editorial Board
  • Author Guidelines
  • Submit your Paper
  • Current Issue
  • Archives
  • Indexing
  • Fees/ APC
  • Reviewers
  • Apply as a Reviewer

IJARAI

  • About the Journal
  • Archives
  • Indexing & Archiving

Special Issues

  • Home
  • Archives
  • Proposals
  • Guest Editors
  • SUSAI-EE 2025
  • ICONS-BA 2025
  • IoT-BLOCK 2025

Future of Information and Communication Conference (FICC)

  • Home
  • Call for Papers
  • Submit your Paper/Poster
  • Register
  • Venue
  • Contact

Computing Conference

  • Home
  • Call for Papers
  • Submit your Paper/Poster
  • Register
  • Venue
  • Contact

Intelligent Systems Conference (IntelliSys)

  • Home
  • Call for Papers
  • Submit your Paper/Poster
  • Register
  • Venue
  • Contact

Future Technologies Conference (FTC)

  • Home
  • Call for Papers
  • Submit your Paper/Poster
  • Register
  • Venue
  • Contact
  • Home
  • Call for Papers
  • Editorial Board
  • Guidelines
  • Submit
  • Current Issue
  • Archives
  • Indexing
  • Fees
  • Reviewers
  • Subscribe

DOI: 10.14569/IJACSA.2025.0160294
PDF

Efficient Personalized Federated Learning Method with Adaptive Differential Privacy and Similarity Model Aggregation

Author 1: Shiqi Mao
Author 2: Fangfang Shan
Author 3: Shuaifeng Li
Author 4: Yanlong Lu
Author 5: Xiaojia Wu

International Journal of Advanced Computer Science and Applications(IJACSA), Volume 16 Issue 2, 2025.

  • Abstract and Keywords
  • How to Cite this Article
  • {} BibTeX Source

Abstract: In recent years, personalized federated learning (PFL) has garnered significant attention due to its potential for safeguarding data privacy while addressing data heterogeneity across clients. However, existing PFL approaches remain vulnerable to privacy breaches, particularly under adversarial inference and client-side data reconstruction attacks. To address these concerns, we propose DP-FedSim, a novel PFL framework incorporating adaptive differential privacy mechanisms. First, to mitigate the limitations posed by fixed-layer personalization strategies, we evaluate parameter significance using the Fisher information matrix. By selectively retaining parameters with higher Fisher values, DP-FedSim reduces the noise impact, enabling more efficient dynamic personalization. Second, we introduce a layered adaptive gradient clipping method. By leveraging the mean and standard deviation of the gradients within each layer, this method allows DP-FedSim to automatically adjust clipping thresholds in response to real-time privacy demands and model states, enhancing the adaptability to various model structures. This ensures a more accurate balance between privacy preservation and model performance. Furthermore, we present a model similarity-based aggregation method utilizing cosine similarity. This technique dynamically adjusts each client's contribution to the global model update, prioritizing clients with models more similar to the global model. This improves the global model's performance and generalization by allowing DP-FedSim to better handle a variety of data distributions and client model attributes. Experimental results on multiple SVHN cifar-10 datasets show that DP-FedSim outperforms the state-of-the-art PFL algorithm by an average of 5% when data heterogeneity is at its strongest. The efficiency of the suggested modules is validated by ablation tests, and the visualization results shed light on the reasoning behind important hyperparameter settings.

Keywords: Federated learning; differential privacy; gradient clipping; model aggregation

Shiqi Mao, Fangfang Shan, Shuaifeng Li, Yanlong Lu and Xiaojia Wu, “Efficient Personalized Federated Learning Method with Adaptive Differential Privacy and Similarity Model Aggregation” International Journal of Advanced Computer Science and Applications(IJACSA), 16(2), 2025. http://dx.doi.org/10.14569/IJACSA.2025.0160294

@article{Mao2025,
title = {Efficient Personalized Federated Learning Method with Adaptive Differential Privacy and Similarity Model Aggregation},
journal = {International Journal of Advanced Computer Science and Applications},
doi = {10.14569/IJACSA.2025.0160294},
url = {http://dx.doi.org/10.14569/IJACSA.2025.0160294},
year = {2025},
publisher = {The Science and Information Organization},
volume = {16},
number = {2},
author = {Shiqi Mao and Fangfang Shan and Shuaifeng Li and Yanlong Lu and Xiaojia Wu}
}



Copyright Statement: This is an open access article licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, even commercially as long as the original work is properly cited.

IJACSA

Upcoming Conferences

IntelliSys 2025

28-29 August 2025

  • Amsterdam, The Netherlands

Future Technologies Conference 2025

6-7 November 2025

  • Munich, Germany

Healthcare Conference 2026

21-22 May 2026

  • Amsterdam, The Netherlands

Computing Conference 2026

9-10 July 2026

  • London, United Kingdom

IntelliSys 2026

3-4 September 2026

  • Amsterdam, The Netherlands

Computer Vision Conference 2026

15-16 October 2026

  • Berlin, Germany
The Science and Information (SAI) Organization
BACK TO TOP

Computer Science Journal

  • About the Journal
  • Call for Papers
  • Submit Paper
  • Indexing

Our Conferences

  • Computing Conference
  • Intelligent Systems Conference
  • Future Technologies Conference
  • Communication Conference

Help & Support

  • Contact Us
  • About Us
  • Terms and Conditions
  • Privacy Policy

© The Science and Information (SAI) Organization Limited. All rights reserved. Registered in England and Wales. Company Number 8933205. thesai.org