Network Traffic Anomaly Detection Based on Deep Q Network

Inhalt:
The new generation of innovative technology has been widely used in all walks of life, so that the dependence of related business on the network has been strengthened, and the importance of network security has become increasingly prominent. The increasingly severe international security situation and the complex and changeable struggle situation also put forward higher requirements for the development speed of cyberspace defense capability and the construction degree of security facilities. As an important part of cyberspace security defense system, the abnormal network traffic detection system is also ushered in a new opportunity under the promotion of intelligent development trend. Based on the application of deep learning in the direction of cyberspace security, this paper chooses anomaly detection as the research goal, and proposes an unsupervised anomaly detection method based on Deep Q-network a single class of the long-term and the short-term memory networks, which can combine the powerful feature extraction ability of deep learning to create a tight envelope around normal data, and help to identify large data sets. Identify and defend potential threats and APT attacks in strong confrontation scenarios in a timely manner, and provide strong support for the improvement of cyberspace security technology system.