Prediction Model Of Network Security Situation Based On Hybrid Optimization RBF Neural Network

The popularity of internet brings us the convenience of life and the transformation of production methods,however it has huge hidden dangers for people's privacy and property security.Therefore combining the historical data and the recent data of network security to predict the future security threats is critical to protecting the health of network environment.This paper studies the relevant theoretical basis of network security situation prediction and the forecasting model which is commonly used,and proposes a network security situation prediction model based on improved artificial fish and particle swarm optimization RBF neural network.The main work is as follows:(1)Through the study of the artificial fish swarm algorithm,it is found that the number of attempts in the foraging behavior affects the convergence efficiency of the algorithm.Therefore,this paper proposes to improve the foraging behavior by using the Metropolis criterion in the simulated annealing algorithm.This method will improve the convergence efficiency and jump out of the local extreme point in the iteration.Besides,the Gaussian mutation operator is introduced into the optimal fish in each iteration of the artificial fish population,and the variance is received by the Metropolis criterion to improve the global search ability of the artificial fish population.(2)By analyzing the advantages and disadvantages of artificial fish swarm algorithm and particle swarm optimization algorithm,the parameters of RBF neural network are optimized by using the global search ability of artificial fish and the local convergence efficiency of particle swarm.According to the characteristics of the sample data,the structure of RBF neural network is designed to construct the network security situation prediction model based on RBF neural network optimized by hybrid algorithm.(3)Based on the simulation experiment and the comparison with other prediction models,the accuracy and performance of RBF neural network prediction model based on hybrid optimization for network security trend prediction are verified.The simulation results show that proposed prediction model is consistent with the actual trend of network security situation,and the prediction accuracy is better than the comparing forecasting model.