A Hybrid Anomaly Detection Model Based on GANomaly in Cloud Environment
Wenkao Yang, Chenglong Zeng · 2022
Anomaly detection technology, which analyzes network data in detail and provides strategies for deploying security tools to ensure the integrity, confidentiality, and reliability of computer systems, is an important part of network information security infrastructure. However, in the cloud environment, the massive network traffic data contains many irrelevant or redundant features, which not only consumes many computing resources but also makes the detection accuracy lower. To this end, this paper proposes a hybrid model based on NSGA-III and GANomaly. Among them, NSGA-III can optimize the model in terms of the number of features, accuracy and false positive rate, while GAN omaly is a classifier that can be trained without negative samples. The experimental results show that the model reduces the number of features by nearly 74% on the basis of an accuracy rate of 99.71 % and a false alarm rate of 0.2% on the CSE-CIC-IDS2018 dataset.