| 入侵检测系统中基于PCA和C-SSGA的双向数据压缩 |
| |
| 引用本文： | 朱永宣,单莘,郭军.入侵检测系统中基于PCA和C-SSGA的双向数据压缩[J].哈尔滨工业大学学报,2009(9):123-127. |
| |
| 作者姓名： | 朱永宣 单莘 郭军 |
| |
| 作者单位： | 北京邮电大学信息工程学院 |
| |
| 基金项目： | 国家自然科学基金资助项目(60475007);教育部跨世纪人才基金资助项目 |
| |
| 摘 要： | 针对入侵检测数据中的冗余特征和冗余实例,提出一种基于主成分分析和混合稳态遗传算法的双向数据压缩方法.利用主成分分析对特征进行压缩,有效地去除特征之间的冗余性;用混合稳态遗传算法进行实例压缩,大大缩减了实例的数量;提出一个基于神经网络的入侵检测系统模型,该模型具有多分类、易于更新系统及快速适应新型入侵的特点.在KDD CUP’99上的实验表明,提出的方法是有效的,可以用于处理大数据集的压缩问题.
|
| 关 键 词： | 入侵检测系统 混合稳态遗传算法 主成分分析 神经网络 |
Bidirectional data reduction based on PCA and combined SSGA in intrusion detection system |
| |
| ZHU Yong-xuan,SHAN Xin,GUO Jun.Bidirectional data reduction based on PCA and combined SSGA in intrusion detection system[J].Journal of Harbin Institute of Technology,2009(9):123-127. |
| |
| Authors: | ZHU Yong-xuan SHAN Xin GUO Jun |
| |
| Affiliation: | (School of Information Engineering,Beijing University of Posts and Telecommunications,Beijing 100876,China) |
| |
| Abstract: | Aimed at the redundant features in intrusion detection data and redundant instances in intrusion detection system,a bidirectional data reduction method based on the principal component analysis(PCA) and combined steady-state genetic algorithm is proposed.Firstly,PCA is employed to reduce the features and the redundant features are removed effectively.Then the combined steady-state genetic algorithm is used to reduce the number of instances.A model of intrusion detection system based on neural networks is presented.The model characterizes itself in muticlassifications and easy updating for the adaptation to new intrusion modes.Experiments on KDD CUP’99 validate the effectiveness of this method.It is indicated that the proposed bidirectional data reduction method can be used to deal with the reduction of big data set. |
| |
| Keywords: | intrusion detection system combined steady-state genetic algorithm(C-SSGA) principal component analysis(PCA) neural networks |
| 本文献已被 CNKI 等数据库收录！ |
