| Summary: | Deep neural networks have demonstrated their effectiveness in most machine learning tasks, with intrusion detection included. Unfortunately, recent research found that deep neural networks are vulnerable to adversarial examples in the image classification domain, i.e., they leave some opportunities for an attacker to fool the networks into misclassification by introducing imperceptible changes to the original pixels in an image. The vulnerability raises some concerns in applying deep neural networks in security-critical areas, such as intrusion detection. In this paper, we investigate the performances of the state-of-the-art attack algorithms against deep learning-based intrusion detection on the NSL-KDD data set. The vulnerabilities of neural networks employed by the intrusion detection systems are experimentally validated. The roles of individual features in generating adversarial examples are explored. Based on our findings, the feasibility and applicability of the attack methodologies are discussed.
|
|---|
