In order to cope with the diversified security challenges faced by non-regulatory business networks, it is particularly important to construct a comprehensive network security protection system. In this paper, a new hybrid kernel function is constructed, the parameters of the support vector machine are optimized by genetic algorithm, and a network security posture indicator system is proposed to realize the network security posture assessment model based on GA-SVM. And the ARIMA model is used to predict the network security posture. The empirical results tabulate that compared with the two prediction models of RBF and PSO-SVM, the accuracy, AUC value, and F1 value of the GA-SVM network security posture assessment model are 89.47%, 0.8792, and 0.8644, respectively, which are able to accomplish the network security assessment in a better way.The results of ARIMA in the task of network security posture prediction have a better fitting effect. Therefore, network security posture assessment and prediction can be used to monitor the security status in the non-regulated business network environment in real time, discover potential threats and abnormal behaviors in a timely manner, and predict possible security risks in the future.
