分类器集成在入侵检测中的应用研究

The application research of classifier ensemble in intrusion detection

针对参与集成的基分类器的选择算法等难点问题,提出一种差异性度量方法以及基于该差异度量进行分类器选择的集成方法.考虑参与集成的基分类器分类准确性和平均差异性,改变最终分类器集合的获取算法,以提高分类器的性能.实验结果表明,此种方法优于bagging方法,能获得更好的检测性能.

Intrusion detection is an important research area of network security. Applying the classifier ensemble to the detection module of an intrusion detection system can improve the detection effect. The application research of classifier ensemble in intrusion detection is studied in this paper, and a diversity measure method and an ensemble method based on the former method are proposed, considering both accuracy and diversity of the base classifiers. Experimental results demonstrate that the ensemble method is better than bagging, and can achieve better detection performance.