基于人工神经网络的主机安全量化评估研究

Research of computer security quantitative evaluation based on artificial neural networks

基于人工神经网络的方法对主机安全性能进行量化评估。分析了BP人工神经网络模型的网络结构及学习算法,分析了影响目标主机安全性能的可能因素,并应用BP神经网络模型对目标主机的安全性能进行样本训练及实际测试。基于人工神经网络的主机安全量化评估为评价目标主机的安全性能提供了可行的方法。

The security performance of target host is assessed quantitatively based on the method of manual neural network. Network framework and learning algorithm of BP neural networks are analyzed and then some factors are brought forward influencing host's security performance. In addition, thetraining samples and practical testing of the host's securitycapability in terms ofBP neural networks are offered. Computer security quantitative assessment based on manual neural network provides a feasible method for the research of assessing security performance of target host.