RESEARCH PAPERIn Progress
Adversarial Examples Against Network Intrusion Detection Systems: A Gradient-Guided Attack Framework
•By Moddux Research
A systematic investigation of adversarial machine learning attacks against network-based intrusion detection systems — demonstrating that gradient-guided perturbations, feature-space manipulation, and black-box transfer attacks can significantly degrade detection accuracy while preserving network functionality. Provides a framework for adversarially robust IDS design.
#Adversarial ML#Intrusion Detection#Machine Learning#Security#Neural Networks#NIDS#Evasion