基于模糊支持向量机的网络入侵检测技术

针对支持向量机算法中存在对噪声数据和异常数据敏感的问题,提出了模糊支持向量机算法,并应用于入侵检测.该算法是在传统支持向量机分类器的构造方法中引入隶属度函数,根据不同输入所得到的分类结果,产生相应的惩罚值.将这个方法应用到入侵检测系统中,能较好地将正常数据和异常数据区分开.实验结果表明,采用模糊支持向量机的入侵检测技术,其误报率低于基于支持向量机的入侵检测,同时其检测率也相对较高.

Network Intrusion Detection Based on Fuzzy Support Vector Classification Machines

A new fuzzy support vector machine is proposed and applied in intrusion detection to salve the sensitivity problem to the noisy data and outlier in the traditional method of SVM(Support Vector Machine).A fuzzy membership function is used in of classification.Fuzzy membership function produces penalty values according to the different affects on the classification result.The method applied in intrusion detection system can divide exactly between normal data and abnormal date.The experimental results show that the Intrusion Detection System(IDS) based on the Fuzzy Support Vector Classification Machines possesses higher detection rate and lower false positive rate than the traditional IDS based on SVM.