基于LM的半自适应遗传神经网络在入侵检测中的应用
首发时间:2010-04-16
摘要:入侵检测是一种主动防御技术,能够实时地对入侵行为进行识别。本文研究了现有的基于遗传神经网络的入侵检测系统及其改进方法,针对传统遗传算法容易早熟和自适应遗传算法计算量大、收敛速度慢等问题,提出了基于LM优化算法的半自适应遗传神经网络。首先,让半自适应遗传算法确定全局最优区域,然后再用LM算法精确搜索。改进后的遗传神经网络充分利用了半自适应遗传算法计算简单且具有全局搜索特性,并结合了LM优化算法局部精确搜索的优点。将该网络应用于入侵检测中,实验结果表明,效果良好。
关键词: 遗传神经网络 LM优化算法 半自适应遗传算法 入侵检测
For information in English, please click here
Application of intrusion detection based on neural network optimized by half adaptive genetic algorithm and LM
Abstract:Intrusion detection is a kind of active defense technology, The thesis has carried on detailed analysis and studies about genetic neural network used in recent years, Aimed at easy premature of the basic genetic algorithm and lots of calculation of adaptive genetic algorithm, the paper put forward half adaptive genetic neural network based on LM optimization algorithm. Firstly, using half adaptive genetic algorithm to determine the global optimal area, then switching to LM algorithm for precise search. Improved genetic neural network makes full use of the half adaptive genetic algorithm with simple calculation and global search, and combining the advantage of local precisely search of LM algorithm. The network is applied to intrusion detection system, experimental results show that the effect is good.
Keywords: genetic neural network LM algorithm half adaptive genetic algorithm intrution detection
基金:
论文图表:
引用
No.4201552632512714****
同行评议
共计0人参与
勘误表
基于LM的半自适应遗传神经网络在入侵检测中的应用
评论
全部评论0/1000