基于δ-HSSVM的协同入侵检测方法

摘要

传统入侵检测方法检测速度较慢.为此,提出一种基于δ-超球结构支持向量机的协同入侵检测方法.增加参数δ,以控制超球内样本的数量,按照网络协议产生3类基于δ-超球结构支持向量机的检测代理,以协同的方式对其相应网络数据做出决策.实验结果表明,与单一检测代理方法相比,该方法检测率高、检测时间少.%For the problem of slow detection speed of traditional intrusion detection, δ-hypersphere Structure Support Vector Machine(δ-HSSVM) is presented. The number of samples inside hypersphere are controled by the increasing parameter S. Three detection agents based on δ-HSSVM generated according to network protocols. And the three detection agents make decisions to the corresponding network data. Experimental results show that, this method has a higher detection rate and less detection time than single detection agent.