In this paper, the method of the hybrid neural network is used to identify malicious TLS traffic, and the network traffic anomaly detection algorithm based on logistic regression and decision tree and the distributed denial of service attack detection algorithm based on hybrid neural network algorithm and gradient lifting tree is proposed. It found that there is a general problem of high feature dimensions in network traffic detection. The useless features will increase the computational complexity of traffic anomaly detection and reduce accuracy. Based on the feature selection method of recursive feature elimination and logistic regression, the importance ranking of flow features obtained. Using the accuracy rate as an indicator, a decision tree algorithm used to model and predict network traffic. During network traffic anomaly detection, it is difficult to distinguish between normal traffic and denial of service attack traffic. The network traffic anomaly detection method based on the decision tree has low computational complexity and can save detection time. At the same time, the accuracy of the algorithm when detecting network traffic anomalies has been significantly improved, and the false alarm rate can also be reduced.
Research on Malicious TLS Traffic Identification Based on Hybrid Neural Network
2020-09-01
1189145 byte
Conference paper
Electronic Resource
English
A Method for TLS Malicious Traffic Identification Based on Machine Learning
Trans Tech Publications | 2021
|Research on Malicious TLS Traffic Detection Based on Spatiotemporal Feature Fusion
Springer Verlag | 2022
|