Multimedia Security Situation Prediction Based on Optimization of Radial Basis Function Neural Network Algorithm

Abstract

Aiming at the problem of prediction accuracy in network situation awareness, a network security situation prediction method based on a generalized radial basis function (RBF) neural network is proposed. This method uses the K-means clustering algorithm to determine the data center and expansion function of the RBF and uses the least-mean-square algorithm to adjust the weights to obtain the nonlinear mapping relationship between the situation value before and after the situation and carry out the situation prediction. Simulation experiments show that this method can obtain situation prediction results more accurately and improve the active security protection of network security. Compared with the PSO-RBF model, AFSA-RBF model, and IAFSA-RBF model, the maximum relative error and minimum relative error of the IAFSA-PSO-RBF model are reduced by 14.27%, 8.91%, and 32.98%, respectively, and the minimum relative error is reduced by 1.69%, 12.97%, and 0.61%, respectively. This shows that the IAFSA-PSO-RBF model has reduced the prediction error interval, and the average relative error is 5%. Compared with the other three models, the accuracy rate is improved by more than 5%, and it has met the requirements for the prediction of the network security situation.