Privacy-preserving federated learning framework with dynamic weight aggregation

Privacy-preserving federated learning framework with dynamic weight aggregation

There are two problems with the privacy-preserving federal learning framework under an unreliable central server.① A fixed weight, typically the size of each participant’s dataset, is used when aggregating distributed learning models on the central server.However, different participants have non-ind...

Full description

Bibliographic Details
Main Author: Zuobin YING, Yichen FANG, Yiwen ZHANG
Format: Article
Language:English
Published: POSTS&TELECOM PRESS Co., LTD 2022-10-01
Series:网络与信息安全学报
Subjects:
federated learning
differential privacy
dynamic aggregation weight
non-independent and identically distributed data
Online Access:https://www.infocomm-journal.com/cjnis/CN/10.11959/j.issn.2096-109x.2022069

Internet

https://www.infocomm-journal.com/cjnis/CN/10.11959/j.issn.2096-109x.2022069

Similar Items