SVM与ANN在网络安全风险评估中的比较研究

Study on the Network Security Risk Evaluation SVM and ANN

该文对SVM和ANN在网络安全风险评估方面的应用进行了比较研究,首先介绍了ANN的工作原理,然后分别将两种方法都应用到网络安全风险评估中,对其评估效果进行了比较。结果发现SVM在小训练样本的分类正确率,泛化能力和训练、测试速度方面,均优越于ANN,是一种更加优越的风险评估方法。

In this paper, SVM and ANN in the network security risk assessment carried out a comparative study of the application, first introduced the working principle of ANN, and then the two methods are respectively applied to the network security risk assessment, their assessment results were compared. The results showed that SVM training samples in a small classification accuracy rate, generalization ability and training, test speed, are superior to ANN, is a more superior risk-assessment methodologies.