| 入侵检测分类算法的研究 |
| |
| 引用本文： | 陈媛媛,杜静. 入侵检测分类算法的研究[J]. 机械工程与自动化, 2006, 0(4): 58-59,62 |
| |
| 作者姓名： | 陈媛媛 杜静 |
| |
| 作者单位： | 1. 中北大学,自动控制系,山西,太原,030051
2. 太原科技大学,经济管理学院,山西,太原,030024 |
| |
| 摘 要： | 分类是数据挖掘的重要技术之一。在入侵检测中用分类算法，如RIPPER程序来构造分类模型。针对RIPPER算法在入侵检测环境下不能很好地对稀有类分类这一缺点，提出了一个改进方案——在RIPPER算法基础上引入两阶段方法，并以DARPA98入侵检测数据为基础做相应实验。实验结果证明改进后的算法对稀有类分类较RIPPER算法准确率高。
|
| 关 键 词： | 入侵检测系统 数据挖掘 分类算法 |
| 文章编号： | 1672-6413（2006）04-0058-02 |
| 收稿时间： | 2006-02-15 |
| 修稿时间： | 2006-02-152006-04-30 |
Study on Classification Algorithm of IDS |
| |
| CHEN Yuan-yuan,DU Jing. Study on Classification Algorithm of IDS[J]. Mechanical Engineering & Automation, 2006, 0(4): 58-59,62 |
| |
| Authors: | CHEN Yuan-yuan DU Jing |
| |
| Affiliation: | 1. Department of Automatic Control of NUC,Taiyuan 030051,China;2. Department of Economy Management of Taiyuan University of Science and Technology,Taiyuan 030024 ,China |
| |
| Abstract: | Classification is a core technique in data mining which can be used to build a classification on model for the sake of distinction between the normal and abnormal in intrusion detection.This paper analyzes the defect that rare classes couldn't be classified by RIPPER and then brings forward the improvement in which the new method of two phases is adopted.At last,the experiment based on data set of DARPA proves that the improved algorithm can effectively classify rare classes. |
| |
| Keywords: | intrusion detection system data mining classification algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
