Abstract: As a kind of active defense technology, intrusion detection system makes up for the shortcomings of traditional security technology. In addition, it also has real-time monitoring function, greatly improves the safety of computer network system. This paper presents a decision rules binary classifier applied for network intrusion detection. The classifier is optimized by a multiobjective evolutionary algorithm in order to maximize both the classification accuracy and the coverage level. The coverage level is the percentage of items that are classified, in opposite to items unable to be classified. Study results provide very good accuracy and coverage level in detecting attacks.

Key words: decision rules, intrusion detection, evolutionary algorithm

摘要： 入侵检测作为一种主动防御技术，弥补了传统安全技术的不足。另外，它还具有实时监测功能，大幅提高了计算机网络系统的安全性能。提出了一种应用于网络入侵检测的决策规则分类器，该分类器对多目标函数的进化算法进行优化，使其分类精度和覆盖率达到最大，其中覆盖率最大就是可分类数据与不可分类数据的比例最大。研究结果表明该分类器对网络攻击有着较好的分类精度和覆盖率。

关键词: 决策规则, 入侵检测, 进化算法