A security situation assessment method based on neural network

Abstract

In the big data environment, the scale of attacks of Distributed Denial of Service (DDoS) continues to expand rapidly. The traditional network situation assessment method cannot effectively evaluate the security situation of DDoS. A security situation assessment method based on deep learning and a security situation assessment model based on neural network are proposed. The model uses convolutional neural network (CNN), back propagation algorithm (BP) and Long Short-Term memory neural network (LSTM) to learn various network security indicators to achieve a comprehensive assessment of the network. The experimental results show that the model can more easily and accurately evaluate the network security status, which is more accurate and flexible than the existing evaluation methods.