Network intrusion detection using generative adversarial networks. (2020)
Type of ContentElectronic Thesis or Dissertation
Thesis DisciplineComputer Science
Degree NameMaster of Engineering
PublisherUniversity of Canterbury
AuthorsZhang, Xiranshow all
Intrusion detection systems (IDS), as one of important security solutions, are used to detect network attacks. With the extensive applications of traditional machine learning algorithms in the security field, intrusion detection methods based on the ma- chine learning techniques have been developed rapidly. However, since the progress of technology and the defects of the intrusion detection system based on machine learning algorithms, the system has gradually failed to meet the requirement for cyber security. Generative Adversarial Networks (GANs) have been widely studied and applied in anomaly detection in recent years thanks to their high potential in learning complex high-dimensional real data distribution. Deep learning techniques can greatly overcome the disadvantages of using traditional machine learning algorithms for intrusion detection. This work proposes to use current existing GANs and their variants for network intrusion detection using real dataset and show the feasibility and comparison results.