ShieldFL Framework Defends Against Byzantine Attacks in Non-IID Federated Learning

Written by

in

Researchers propose the ShieldFL framework, which defends against Byzantine attacks in non-IID settings through K-Means client clustering based on data distribution features, cosine similarity evaluation of normalized gradients, and Medoid selection.

Source:BIOENGINEER.ORG · Read Original

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *