摘要
针对现有入侵检测系统识别率低、误报率高的问题,将SOM神经网络应用到入侵检测系统。自组织特征映射神经网络SOM(Self Organizing Feature Maps)作为一种优良的聚类工具,具有无需监督,能自动对输入模式进行聚类的优点。为验证检测方法的有效性,采用KDDCup99的训练集与测试集进行实验。
For the low identification rate and high false alarm rate of existing Intrusion Detection System,we applied the SOM neural network to the intrusion detection system.As a good clustering tool,SOM can automatically cluster input mode with no supervision.To verify the validity of testing,we use the training set and test set of KDD Cup99 to make experiments.
出处
《电脑知识与技术》
2010年第7X期5711-5713,共3页
Computer Knowledge and Technology
关键词
自组织特征映射
聚类
入侵检测
self organizing feature maps
clustering
intrusion detection
