An Efficient Improvement of CMA-ES Algorithm for the Network Security Situation Prediction

ABSTRACT

An improved covariance matrix adaptation evolution strategy algorithm (CMA-ES) is proposed and it is used to train the forecasting model of the network security situation in this paper. A new recombination strategy which adds a heuristic component is developed in the improved CMA-ES algorithm, and the search speed is accelerated. The experimental results show that, compare with original algorithm and its variants, the improved CMA-ES algorithm can greatly increased the search speed in high dimensional problems. The improved CMA-ES algorithm is an efficient evolutionary algorithm which can be applied to the network security situation prediction.

Keywords:

Network security situation prediction, covariance matrix adaptation evolution strategy, evolutionary algorithm.