Improved Population Intelligence Algorithm and BP Neural Network for Network Security Posture Prediction

Abstract

To address the problems of low prediction accuracy and slow convergence of the network security posture prediction model, a population intelligence optimization algorithm is proposed to improve the network security posture prediction model of the BP neural network. First, the adaptive adjustment of the two parameters with the increase of iterations is achieved by improving the inertia weights and learning factors in the particle swarm optimization (PSO) algorithm so that the PSO has a large search range and high speed at the initial stage and a strong and stable convergence capability at the later stage. Secondly, to address the problem that PSO is prone to fall into a local optimum, the genetic operator is embedded into the operation process of the particle swarm algorithm, and the excellent global optimization performance of the genetic algorithm is used to open up the spatial vision of the particle population, revive the stagnant particles, accelerate the update amplitude of the algorithm, and achieve the purpose of improving the premature problem. Finally, the improved algorithm is combined with the BP neural network to optimize the BP neural network and applied to the network security posture assessment. The experimental comparison of different optimization algorithms is applied, and the results show that the network security posture prediction method of this model has the smallest error, the highest accuracy, and the fastest convergence, and can effectively predict future changes in network security posture.