Zero-Day Attack Detection with Unsupervised Anomaly Detection

EasyChair Preprint 13985

Abstract

Zero-day attacks pose a significant threat to the security of computer systems, as they exploit unknown vulnerabilities that have not been addressed by security patches. Traditional signature-based detection methods are ineffective against these attacks, as they rely on known patterns. This paper proposes a novel approach to zero-day attack detection using unsupervised anomaly detection techniques. By analyzing the behavior of network traffic, our system can identify anomalous patterns that may indicate the presence of a zero-day attack. We evaluate the effectiveness of our approach using a real-world dataset and demonstrate its ability to accurately detect zero-day attacks with low false positive rates. The proposed method provides a promising solution for early detection and mitigation of zero-day attacks, enhancing the overall security posture of computer systems. Further research is needed to refine and improve the performance of the system, as well as to explore its application in other domains.

Keyphrases: Attacks, Mitigation, zero-day

@booklet{EasyChair:13985,
  author    = {Ralph Shad and Ayoolu Olukemi and Axel Egon},
  title     = {Zero-Day Attack Detection with Unsupervised Anomaly Detection},
  howpublished = {EasyChair Preprint 13985},
  year      = {EasyChair, 2024}}