About this document

Presentation 3 by e.umaramesh is a document available to read on EtoBox.

The document presents a proactive privacy-preserving defense for a Federated Learning-based Intrusion Detection System (FL-IDS) that addresses vulnerabilities such as gradient leakage and model poisoning. It outlines a hybrid CNN-LSTM architecture utilizing Packed Homomorphic Encryption and a proactive defense mechanism to detect and mitigate malicious client updates while maintaining data privacy. The proposed system includes a recovery module to remove harmful contributions from the global model, aiming t

Author

e.umaramesh

Language

EN