摘要： 为改善入侵检测系统的性能，提出一种基于知识约简的特征提取方法，根据粗糙集理论给出入侵检测系统的形式化描述，使用知识约简提取属性特征，通过信息损耗和信息增益分别控制连续数值属性特征的离散化和属性特征的约简过程。实验结果证明，该方法可有效消除初始数据中的冗余信息和数据噪声。

关键词: 入侵检测, 粗糙集, 属性约简, 特征提取

Abstract: In order to improve the performance of Intrusion Detection System(IDS), this paper proposes a feature extraction method based on knowledge reduction. Rough set theory is used to do the formal description for IDS. Knowledge reduction is used to extract attribute features. Information loss and information gain are individually used to control the discrete procedure of continuous value attributes and the reduction of attribute features. Experimental result justifies that the method can eliminate the redundant information and noise of initial data effectively.

Key words: intrusion detection, rough set, attribute reduction, feature extraction

中图分类号: 

• TP309.2