自适应遗传算法优化神经网络的入侵检测研究  

Research of intrusion detection based on neural network optimized by adaptive genetic algorithm

作　　者：栾庆林[1] 卢辉斌[1] 

机构地区：[1]燕山大学信息科学与工程学院,河北秦皇岛066004

出　　处：《计算机工程与设计》2008年第12期3022-3025,共4页Computer Engineering and Design

摘　　要：入侵检测是一种动态的安全防护技术,能够对网络内部、外部攻击进行防御。基于神经网络的入侵检测是常见的智能入侵检测方法。针对神经网络算法易陷入局部极值和简单遗传算法收敛速度慢的问题,提出了一种将神经网络和遗传算法相结合,用遗传算法优化神经网络权值,在遗传算法优化神经网络时采用自适应遗传操作。将自适应遗传算法优化神经网络算法应用于入侵检测系统中,实验结果表明,该方法能够有效的提高系统的检测率,降低误报率和漏报率。The intrusive detection technology is a dynamic security protection technology, it can detect the interior and exterior attacks. The research of intrusion detection based on the neural network is the familiar way of intelligent intrusion detection. Aimed at the problems that neural network is prone to the local optimum and simple genetic algorithm has the shortcoming of slow convergence, it shows a way that combines neural network with genetic algorithm. It makes use of genetic algorithm to optimize the weights of the neural network. Adaptive genetic operation is presented in the course of optimization. And the algorithm is applied into the intrusion detection systems, the result of the experiments shows the method is good for the improvement of the detection rate, and reduction ofmisreporting rate.

关 键 词：入侵检测 神经网络 遗传算法 自适应遗传算法 权值优化 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]