ホーム
探索
nav.journalClub
トレンド
その他
synapse
⌘+K
言語
日本語
日本語
FedDRLPD: Deep reinforcement Learning-Based defense mechanism against poisoning attacks in federated learning | Synapse
March 3, 2026
FedDRLPD: Deep reinforcement Learning-Based defense mechanism against poisoning attacks in federated learning
NX
Nuo Xu
YF
Yong Feng
NL
Nianbo Liu
University of Electronic Science and Technology of China
See all
Key Points
Evidence suggests the implementation of deep reinforcement learning significantly reduces poisoning attacks.
The analysis reported a reduction of over 30% in successful attacks when using the proposed mechanism.
Assessment using simulations demonstrated the effectiveness of the federated learning framework and the deep reinforcement learning model.
The findings highlight the need for advanced security protocols in federated learning to counteract ongoing threats.
Mark Helpful
Like
Save
Bookmark
Relay
Share
Mark Helpful
Like
Save
Bookmark
Relay
Share
Cite This Study
Copy
Xu et al. (Sat,) studied this question.
synapsesocial.com/papers/69a76143c6e9836116a2f09d
https://doi.org/https://doi.org/10.1016/j.knosys.2026.115558