| 基于数据挖掘的入侵检测多分类模型研究 |
| |
| 引用本文： | 沈利香,曹国.基于数据挖掘的入侵检测多分类模型研究[J].常州工学院学报,2011(6):36-40. |
| |
| 作者姓名： | 沈利香 曹国 |
| |
| 作者单位： | 常州工学院计算机信息工程学院;常州工学院经济与管理学院 |
| |
| 摘 要： | 入侵检测系统捕获的大量网络数据,需要采取有效的算法进行分类,用以判别数据流是否异常以及所属的攻击类别。为了有效地挖掘出判别数据流的模式与规则,针对入侵检测系统的分类器模型,采用数据挖掘技术中的反向传播神经网络算法、有监督的Kohonen神经网络算法和支持向量机算法进行对比研究。主要分析了各个算法的分类正确率、检测率和误...
|
| 关 键 词： | 支持向量机 反向传播神经网络 有监督的Kohonen 入侵检测系统 |
Research of an Intrusion Detection Multi-classification Model Based on Data Mining |
| |
| SHEN Li-xiang,CAO Guo.Research of an Intrusion Detection Multi-classification Model Based on Data Mining[J].Journal of Changzhou Institute of Technology,2011(6):36-40. |
| |
| Authors: | SHEN Li-xiang CAO Guo |
| |
| Affiliation: | 1.School of Computer & Information Engineering,Changzhou Institute of Technology,Changzhou 213002; 2.School of Economics & Management,Changzhou Institute of Technology,Changzhou 213002) |
| |
| Abstract: | Intrusion detection system captures a large amount of network data,which should be classified by effective algorithms to identify network data streams.To find out the models and rules used by IDS′ classifier,data mining technologies are used to carry out comparative research,including Back Propagation Neural Network(BPNN),supervised-Kohonen Neural Network and Support Vector Machine(SVM) to make main analysis of classification accuracy,detection rate and false alarm rate.The experimental results carried on the KDD99 data set show that each algorithm has some advantages.SVM tends to be better,concerning the consideration of comprehensive factors. |
| |
| Keywords: | SVM BPNN supervised-Kohonen IDS |
| 本文献已被 CNKI 维普 等数据库收录！ |
