EvoIDS is a self-evolving intrusion detection system that detects concept drift, adapts to changing network traffic patterns, and identifies zero-day attacks using a multi-detector open-set recognition framework.
autoencoder lightgbm concept-drift intrusion-detection-system network-intrusion-detection cloud-security isolation-forest continual-learning drift-detection open-set-recognition cicids2017 zero-day-detection adaptive-machine-learning
-
Updated
Jun 9, 2026 - Jupyter Notebook