• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

计算机工程与科学

• 人工智能与数据挖掘 • 上一篇    下一篇

一种有效的Gk-prototypes聚类算法

郭映江,徐蔚鸿,陈沅涛,文泽林   

  1. (长沙理工大学计算机与通信工程学院,湖南 长沙 410114)
  • 收稿日期:2018-09-03 修回日期:2019-01-03 出版日期:2019-09-25 发布日期:2019-09-25
  • 基金资助:

    国家自然科学基金(61702052);湖南省科技服务平台专项(2012TP1001);湖南省教育厅重点项目(17A007);综合交通运输大数据智能处理湖南省重点实验室项目(2015TP1005);长沙市科技计划项目(KQ1703018,KQ1706064)

A novel Gk-prototypes clustering algorithm

GUO Ying-jiang,XU Wei-hong,CHEN Yuan-tao,WEN Ze-lin   

  1. (School of Computer and Communication Engineering,Changsha University of Science & Technology,Changsha 410114,China)
     
  • Received:2018-09-03 Revised:2019-01-03 Online:2019-09-25 Published:2019-09-25

摘要:

针对传统的聚类算法对初始聚类中心敏感、只能对单一属性聚类且聚类效果有时欠佳等不足,提出了一种能处理数值属性和分类属性的Gk-prototypes聚类算法。
在经典的k-prototypes聚类算法的基础上,利用去模糊相似矩阵来构造粗粒子集,结合粒计算和最大最小距离法确定初始聚类中心,并改进了目标函数。实验结果和理论分析表明,Gk-prototypes聚类算法与其他基于k-prototypes的改进算法相比,聚类更准确,有效性更好,鲁棒性更强。
 

关键词: k-prototypes聚类, 去模糊相似矩阵, 粒计算, 最大最小距离法

Abstract:

Traditional clustering algorithms are sensitive to initial clustering centers, and their clustering effect is sometimes poor. For these reasons, we present a Gk-prototypes clustering algorithm to process numerical properties and classification properties. Based on the classical k-prototypes clustering algorithm, the proposed algorithm uses the de-fuzzy similarity matrix to construct coarse particle sets, and employs particle calculation and the maximum and minimum distance method to determine the initial clustering center, thus the objective function is improved. Experimental results and theoretical analysis show that the Gk-prototypes clustering algorithm is more accurate, more effective and more robust than other improved algorithms based on k-prototypes.
 

Key words: k-prototypes clustering, de-fuzzy similarity matrix, granular computing, maximum minimum distance method