| 一种新的基于粗糙集的leader聚类算法 |
| |
| 引用本文： | 张琼,张莹,白清源,谢丽聪,谢伙生.一种新的基于粗糙集的leader聚类算法[J].计算机科学,2008,35(3):177-179. |
| |
| 作者姓名： | 张琼 张莹 白清源 谢丽聪 谢伙生 |
| |
| 作者单位： | 福州大学数学与计算机科学学院,福州,350002 |
| |
| 基金项目： | 福州大学校科研和教改项目
,
福建省教育厅科研项目 |
| |
| 摘 要： | 传统聚类方法将对象严格地划分到某一类,但很多时候边界对象不能被严格地划分.粗糙集用上近似集和下近似集表示一个类,对这种边界不确定的处理非常有效,典型算法有基于粗糙集的k-means聚类算法和基于粗糙集的leader聚类算法.本文针对RFA(R0ughFuzzyApproach)算法存在的不足,提出了一种新的基于粗糙集的leader聚类算法(NRL,Novel Rough-based Leader).其基本思想是首先数据项由于与其最近类中心的距离不同,分别被划分到leader集或者supporting leader集,然后对leader集和supporting leader集进行标号,得到聚类结果.实验结果表明NRL算法非常有效.
|
| 关 键 词： | 聚类 粗糙集 k-means算法 leader算法 |
A Novel Rough-based Leader Clustering Algorithm |
| |
| ZHANG Qiong,ZHANG Ying,BAI Qing-Yuan,XIE Li-Cong,XIE Huo-Sheng.A Novel Rough-based Leader Clustering Algorithm[J].Computer Science,2008,35(3):177-179. |
| |
| Authors: | ZHANG Qiong ZHANG Ying BAI Qing-Yuan XIE Li-Cong XIE Huo-Sheng |
| |
| Abstract: | Objects are partitioned into clusters with crisp boundaries in the conventional algorithms. However,clusters do not necessarily have crisp boundaries. Rough set is represented with lower-bound and upper-bound,and is good for the case. At present,there have been some typical algorithms,such as the rough-based k-means clustering algorithm and the rough-based leader clustering algorithm. In this paper,a novel rough-based leader clustering algorithm is proposed,since there are some disadvantages in the RFA algo... |
| |
| Keywords: | Clustering Rough set K-means algorithm Leader algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
点击此处可从《计算机科学》下载全文 |
|
