Research Of Network Security Situation Assessment And Prediction Method

With the advent of the information age, the Internet develops rapidly and becomes popular. But at the same time network security incidents occurred frequently, which brought serious harm to people's life.Although there are a variety of network security devices used to protect the security of the Internet. But due to the specificity and different emphasis, they are often designed only for a particular security problems,which results to they can't make a good evolution and prediction of the overall network security status.In this paper, through the detailed analysis and research of the current network security situation assessment and prediction method, it puts forward with the network security predict method that uses the Bisect K-means clustering algorithm to determine the data center to improve the radial basis function neural network, which makes up for the difficulty of determine the data center. This method is verified by experiments and improves. the prediction accuracy under some certain conditions.The paper also puts forward with the network security situation predict method that uses equivalent interval partitioning strategy to improve the artificial immune algorithm, which makes up for defect of the data redundancy in the process of generating the initial antibody population. This method is verified by experiments and improves the convergence speed.