基于核表示的协同入侵检测方法

摘要

Many classification methods are very difficult to classify in the low dimensional data,but it is extended to higher dimensional spaces through the kernel function,which overcome the limitations of the classification of low dimensional data,and can be directly used in intrusion detection.The kernel representational method can be as a nonlinear extension of conventional representational method.The proposed method includes data collectors,data preprocessing,detection module and disposing module.The results of experimental comparison show that the collaborative method has a higher detection rate and lower time complexity.%许多分类方法在低维数据上难以进行分类,但通过核函数扩展到高维空间,可以克服低维数据分类的局限性,并可直接运用于入侵检测.提出了一种基于核表示的协同入侵检测方法,该方法可以看作是传统的基于表示方法的非线性扩展.协同入侵检测方法的主要模块有数据收集器、数据预处理、检测模块和处理模块.实验对比结果表明,提出的协同入侵检测方法具有较高的检测率和较低的时间复杂度.