基于PCA-SVM的信息安全风险评估模型

Assessment model of information security risk based on PCA-SVM

提出了一种基于主成分分析(PCA)和支持向量机(SVM)的信息安全风险评估模型。首先运用层次分析法构建信息安全风险评估指标体系并采用主成分分析法对风险影响因素进行降维;接着将主成分作为SVM学习样本的输入向量;并利用粒子群算法优化支持向量机的惩罚系数C和核宽度系统σ,建立了一种智能化的信息安全风险评估模型。仿真结果表明,PCASVM方法与标准SVM和BP神经网络相比,有较高的分类准确率,是一种优异的信息安全风险评估模型。

Assessment model of information security risk（ ISRA） is established in this paper based on principal component analysis and support vector machine（ PCA-SVM）. First,ISRA index system is established by analytic hierarchy process（ AHP）. Then,new comprehensive indexes are generated by principle component analysis. Third,support vector machine is trained with principal component as the input vectors. Fourth,PSO algorithm is used to optimize the SVM＇s penalty factor C and kernel width coefficient σ. The simulation results suggested that, the proposed PSO-SVM method has higher classification accuracy than the BP neural network method,PSO-BP method and the standard SVM method.