Contact: Alexander Hartl, Maximilian Bachl, Fares Meghdouri. Explainability methods and Adversarial Robustness metrics for RNNs for Intrusion Detection Systems. Also contains code for "SparseIDS: Learning Packet Sampling with Reinforcement Learning" (branch "rl").

Toolbox for analysis of model's quality and model's description. For further details see

This repository contains the proposed Federated Feature Unlearning framework, which introduces feature-level unlearning in federated environments by utilizing local feature sensitivity.