基于加权多随机决策树的入侵检测模型  

Intrusion detection model based on weighted multi-random decision tree

作　　者：赵晓峰[1] 叶震[1] 

机构地区：[1]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《计算机应用》2007年第5期1041-1043,共3页journal of Computer Applications

摘　　要：传统的决策树分类方法(如ID3和C4.5)对于相对小的数据集是很有效的。但是,当这些算法用于入侵检测这样的非常大的数据时,其有效性就显得不足。采用了一种基于随机模型的决策树算法,在保证分类准确率的基础上,减少了对系统资源的占用,并设计了基于此算法的分布式入侵检测模型。最后通过对比试验表明该模型在对计算机入侵数据的分类上有着出色的表现。The traditional decision tree category methods （such as＂ ID3, CA. 5） are effective on small data sets. However, when these methods are applied to massive data of IDS, its effectivity will get influenced. In this paper, a random model based decision tree algorithm was applied, and an intrusion detection model based on it was provided. It is verified by experiment that this model is evidently powerful for IDS.

关 键 词：决策树 入侵检测 分辨矩阵 随机决策树 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]