摘要
本文对传统的粗糙神经网络模型进行了改进,加入了具有不良信息过滤功能的隐单元,在此基础上提出了相应的网络入侵检测模型,充分发挥了粗糙集理论和神经网络的优势,弥补了各自的缺点。实验证明,在传统神经网络里加入具有不良信息过滤功能的隐单元,可以有效识别网络中的不良信息,降低神经网络系统的输入维度,提高入侵检测系统的识别效果。
In this paper, the traditional rough neural network model is improved, adding a hiding unit with function of filtrating harms information. Then an intrusion detection model is proposed to show the advantages of both rough sets and neural network. With running an example, this model can identify harm information in WebPages and improve the efficiency of intrusion detection system obviously.
出处
《计算机科学》
CSCD
北大核心
2008年第4期63-65,共3页
Computer Science
关键词
粗糙神经网络
入侵检测
不良信息过滤
模型
Rough neural network, Intrusion detection,Filtrating harm information,Model
