计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (15): 118-120.

• 网络、通信、安全 • 上一篇    下一篇

基于Boosting RBF神经网络的入侵检测

党长青,刘 洁,牛分中   

  1. 唐山学院 信息工程二系,河北 唐山 063020
  • 收稿日期:2007-09-05 修回日期:2007-11-26 出版日期:2008-05-21 发布日期:2008-05-21
  • 通讯作者: 党长青

Intrusion detection based on boosting method and RBF neural network

DANG Chang-qing,LIU Jie,NIU Fen-zhong   

  1. Department Two of Information Engineering,Tangshan College,Tangshan,Hebei 063020,China
  • Received:2007-09-05 Revised:2007-11-26 Online:2008-05-21 Published:2008-05-21
  • Contact: DANG Chang-qing

摘要: 提出一种新颖的基于boosting RBF神经网络的入侵检测方法。将模糊聚类和神经网络技术相结合,提出基于改进的FCM算法和OLS算法相结合的FORBF算法,为了提高RBF神经网络的泛化能力,采用Boosting方法,进行网络集成。以“KDD Cup 1999 Data”网络连接数据集训练神经网络并仿真实验,得到了较高的检测率和较低的误警率。

关键词: 入侵检测, Boosting方法, RBF神经网络, 正交最小二乘法

Abstract: A novel method is proposed for intrusion detection based on improved boosting RBF neural network.By combining fuzzy clustering and neural network,this paper puts forward an FORBF algorithm based on improved FCM and OLS.In order to improve the precision of the RBF neural network for intrusion detection,the improved boosting algorithm is used to build an integration-neural network.Simulated experiments with KDD Cup 1999 network connections data have shown that the proposed method is effective for intrusion detection owing to excellent performance of the higher attack detection rate with lower false positive rate.

Key words: intrusion detection, Boosting method, RBF neural network, orthogonal least squares