作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2008, Vol. 34 ›› Issue (15): 168-169,. doi: 10.3969/j.issn.1000-3428.2008.15.061

• 安全技术 • 上一篇    下一篇

基于MPSO的BP网络及其在入侵检测中的应用

肖晓丽,黄继红,刘志朋   

  1. (长沙理工大学计算机与通信工程学院,长沙 410076)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2008-08-05 发布日期:2008-08-05

BP Networks Based on MPSO and Its Application in Intrusion Detection

XIAO Xiao-li, HUANG Ji-hong, LIU Zhi-peng   

  1. (College of Computer & Communication Engineering, Changsha University of Science and Technology, Changsha 410076 )
  • Received:1900-01-01 Revised:1900-01-01 Online:2008-08-05 Published:2008-08-05

摘要: 提出一种基于变异粒子群优化(MPSO)的BP网络学习算法,该算法用PSO算法替代了传统BP算法,且在学习过程中,引入变异操作,克服传统BP算法易陷入局部极小和PSO算法早熟的不足。并把该算法应用于入侵检测中,通过KDD99 CUP数据集分别对基于不同算法的BP神经网络进行了仿真实验比较,结果表明,该算法的收敛速度快,迭代次数较少,而且测试平均准确率高达96.5%。

关键词: 粒子群优化算法, 遗传算法, BP神经网络, 入侵检测, 变异

Abstract: This paper presents a BP neural networks learning algorithm based on Mutation Particle Swam Optimization(MPSO) algorithm. The MPSO algorithm substitutes the traditional BP algorithm during the learning process in order to overcome that the traditional BP algorithm is easy to trap local minima and the PSO algorithm is precocious. The algorithm is applied to intrusion detection. The diagnostic results among BP neural networks based on different algorithms are compared by using KDD999 CPU dataset. The conclusion is that BP neural networks based on MPSO has faster convergence rate, minimum iterations, higher accuracy.

Key words: Particle Swam Optimization(PSO) algorithm, genetic algorithm, BP neural networks, intrusion detection, mutation

中图分类号: