摘要： 提出一种基于Boosting算法的入侵检测方法。先用神经网络初步确定一个入侵检测函数,在此基础上,利用Boosting方法构造一个基于神经网络的入侵检测函数序列,然后以一定的方式将它们组合成一个加强的总检测函数,据此进行入侵检测。实验结果显示,这种方法明显提高了检测性能。

关键词: 入侵检测, 神经网络, Boosting算法

Abstract: This paper presents a new algorithm based on Boosting for intrusion detection. Firstly, it builds a function for intrusion detection using neural network, then uses boosting algorithm and builds a serials of function based on the neural network. At last, a final function based on boosted combination of those functions is found, which is applied in intrusion detection. Experimen tresults show good detective performance of the algorithm.

中图分类号: 

• TP393.08