基于遗传神经网络的入侵检测模型

The Intrusion Detection Model Based on Genetic Neural Network

针对入侵检测系统中存在的对入侵事件高误报率和漏报率问题,提出了遗传神经网络,该方法基于遗传算法的全局搜索和BP网络局部精确搜索的特性,利用遗传算法优化网络初始权重,将遗传算法和BP算法有机结合.实验结果表明,该算法正确鉴定合法的用户矢量为93%,发生7%的误报率.与BP、GA算法相比,分别高出2.875%和5.562%.

To solve the problem of high rate of false negatives and false positives of IDS, the genetic neural network is proposed. This method, based on the traits that the genetic algorithms (GA) are good in global searching, and the back propagation ( BP) are effective on accurate local searching and joining the genetics algorithm and BP algorithm together and optimizing the initial weights of BP with GA. The experiments proved that this kind of method has 93% detection rate, 7% false negatives, compared with BP and GA, higher 2. 875% and 5. 562% respectively.