论文部分内容阅读
针对煤炭网络入侵检测数据属性非线性冗余的特点,提出了一种核主成分分析支持向量机检测方法。该方法可以有效地对非线性冗余属性提取主成分,提高了入侵检测的精度。在kdd cup 99数据集上的实验表明本文方法较传统支持向量机检测方法和主成分分析支持向量机检测方法具有明显的优势。
In view of the characteristics of coal network intrusion detection data non-linear redundancy, a kernel principal component analysis support vector machine detection method is proposed. This method can effectively extract the principal components of nonlinear redundancy attributes and improve the accuracy of intrusion detection. Experiments on the kdd cup 99 dataset show that the proposed method has obvious advantages over traditional support vector machine (SVM) detection and principal component analysis support vector machine (SVM) detection methods.