基于粗糙集和神经网络理论的入侵检测分类器的设计与实现  

Design and Implementation of Intrusion Detection Classifier Based on Rough Sets and Neural Network

作　　者：张小松[1] 郭琳虹[1] 窦炳琳[2] 

机构地区：[1]唐山学院信息工程系,河北唐山063000 [2]华北电力大学信息与网络管理中心,河北保定071003

出　　处：《唐山学院学报》2007年第2期18-20,共3页Journal of Tangshan University

摘　　要：入侵检测系统中分类器的设计是整个系统中重要部分。提出一种基于粗糙集约简算法的神经网络分类器模型,将粗糙集能够在保持系统精度的情况下去掉冗余属性引入到神经网络中,优化网络结构。实验测试表明,此模型能够保证系统检测的准确性,具有一定的实用性。The design of classifier is the most important part in intrusion detection system. A new neural network classifier model is proposed based on rough sets and the structure is optimized by the use of rough sets in neural network that can get rid of the redundant attributes by still maintaining the precision of intrusion detection system. The result shows that the model can guaranttee the veracity of system detection and has some practicability.

关 键 词：入侵检测 分类器 粗糙集 神经网络 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]