Personalised federated learning based on probabilistic knowledge transfer

Mengwei Yan; Longfei Han

doi:10.1117/12.3031025

19 July 2024 Personalised federated learning based on probabilistic knowledge transfer

Mengwei Yan, Longfei Han

Author Affiliations +

Proceedings Volume 13181, Third International Conference on Electronic Information Engineering, Big Data, and Computer Technology (EIBDCT 2024); 131818E (2024) https://doi.org/10.1117/12.3031025
Event: Third International Conference on Electronic Information Engineering, Big Data, and Computer Technology (EIBDCT 2024), 2024, Beijing, China

Abstract

Federated learning is mainly to solve the problem of data privacy in distributed machine learning, but the current non independent and identically distributed (non-iid) data poses a serious challenge to federated learning, which not only destroys the performance of the global model, but also affects the personalized features of the local model more importantly. In this paper, we try to keep local personalization by an idea of adding a private model locally for knowledge transfer to the local model, in addition, we also combine a kind of matching the probability distributions of the data of the two models in the feature space to do the knowledge transfer, which can improve the effect of personalization of the local model as well as improve the performance of the global model more effectively.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Mengwei Yan and Longfei Han "Personalised federated learning based on probabilistic knowledge transfer", Proc. SPIE 13181, Third International Conference on Electronic Information Engineering, Big Data, and Computer Technology (EIBDCT 2024), 131818E (19 July 2024); https://doi.org/10.1117/12.3031025

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE