摘要
针对数据集中若存在孤立点或者是噪声数据会影响模糊C均值聚类算法(FCM)的聚类性能问题,本文将离群点的辨认方法与FCM算法相结合,提出一种改进的FCM聚类算法。该算法有效地降低了孤立点或噪声数据对正常数据的影响,提高了FCM算法的聚类精度。将该算法在入侵检测系统中进行实验验证,通过与FCM算法进行对比分析,证明了该算法的有效性和可行性。
Concerning the influence of the isolation data or noise to FCM,a new kind of method proposed in this paper.It's a improved fuzzy C-means clustering algorithm based on outlier identification.Proposed algorithm eliminated the influence of the isolation data or noise to normal datas and improved the clustering capacity of the FCM.Through comparatived and analysied with FCM algorithm tests show that the algorithm has good effectiveness and feasibility when it applied to intrusion detection.
出处
《微计算机信息》
2012年第1期161-162,172,共3页
Control & Automation
基金
湖南省科技计划资助项目(2009GK3012)
湖南省自然科学基金项目(09JJ3120)
关键词
模糊聚类
模糊C均值聚类算法
离群点辨认
入侵检测
fuzzy clustering
fuzzy C-means clustering algorithm
outlier identification
intrusion detection