摘要
提出一种基于约束模糊聚类思维的网络入侵检测方法,首先对疑似入侵数据进行预处理,将网络数据分割为若干区域,将数据分配至对应的区域中,以区域为单位实现分类;通过区域进化法,以模糊聚类的思想为约束,利用文档中类标签信息引导区域的进化过程,用得到的结果对网络中的未知数据进行迭代分类,以实现网络入侵检测.实验结果表明,与传统方法相比所提方法检测精度高,所需时间更短.
In this paper, a new method of thinking on a const raint fuzzy cluster ing to detect the network intrusion is proposed. Fi rst, wi thin the preprocessing of al l suspected int rusion data, the network data are divided into a number of regions, and then al located to the cor responding data area to realize classif icat ion in each region. Through the method of regional evolut ion, and restrained by a fuzzy clustering thinking, we are able to apply the resul t to the i terat ive classif icat ion on the unknown data in the network whi le using the simi lar tag information in the documents to guide the evolut ion of the region. Thus the network int rusion will be detected. The exper iment shows that we are able to get higher precision in a shorter period af ter adopt ing the proposed method.
出处
《湘潭大学自然科学学报》
北大核心
2017年第3期61-64,84,共5页
Natural Science Journal of Xiangtan University
基金
辽宁省教育厅项目(JG14EB097
JG16EB089)
关键词
入侵检测
模糊聚类
约束
区域进化
int rusion detect ion
fuzzy cluster ing
const raints
regional evolut ion
