Network Information Security Technology and Application Based on SOM-BP Neural Network

Network Information Security Technology and Application Based on SOM-BP Neural Network

연구 분야: Networking

학회: 2024 IEEE 13th International Conference on Communication Systems and Network Technologies (CSNT)

With the opening up of network system, the number of power companies continues to grow in the securities markets. On the basis of summarizing the previously studied model algorithms and taking into account the special market environment, this paper extracts the most essential core of studies on network information security early-warning and proposes a network information security system for power enterprises by models of information security. Finally, the network information security early-warning and monitoring system is built through BP neural network using the classified five grades of samples. In addition, predictive accuracy of the proposed model algorithm is analyzed by comparison with other model algorithms. The results show that the proposed network information security early-warning system can further classify the relatively optimal ST and non-ST sample data obtained into five grades as the input information of SOM network, namely crisis, severe warning, warning, alert and normal, to finally acquire five grades of sample objects.