基于支持向量机的入侵检测系统的研究

Research of Intrusion Detection System Based on Support Vector Machine

对入侵检测和支持向量机的知识进行了基本的介绍,概述了支持向量机实现入侵检测的基本思想,提出了一个基于支持向量机的入侵检测模型,并对其中各个模块进行功能介绍,然后将支持向量机引入到入侵检测系统中。利用KDD99入侵检测数据进行了仿真实验,分析了该模型的工作过程。实验结果表明:该模型避免了高维特征空间的复杂计算,较好地解决了小样本、非线性、高维数、局部极小点等实际问题,能够较好地检测出入侵行为。

Introduces the concepts about the intrusion detection system and support vector machine, summarizes the basic idea about the support vector machine acjoeving intrusion detection, presents a total intrusion detection model based on SVM, and analyses the function of the each model in the frame. The support vector machine is induced into the intrusion detection system. The emulational experiment was done by using KDD99 data sets, analysed the working process of the model. The experiment results demonstrate that the model avoid complex calculation in the high dimension, can detect the intrusion. It resolves the pratical,cal problem about the small number of the training data, nonlinear, high dimension and the local smallest point.