电子科技 ›› 2019, Vol. 32 ›› Issue (11): 70-73.doi: 10.16180/j.cnki.issn1007-7820.2019.11.014

• • 上一篇    下一篇

一种结合贪心选择和特征加权的高维数据聚类算法

向志华,邵亚丽   

  1. 广东理工学院 信息技术学院,广东 肇庆 526100
  • 收稿日期:2017-11-30 出版日期:2019-11-15 发布日期:2019-11-15
  • 作者简介:向志华(1982-),女,讲师。研究方向:机器学习与数据挖掘。
  • 基金资助:
    广东省教育厅科技项目(201713720010)

A High Dimensional Data Clustering Algorithm Combining Greedy Selection and Feature Weighting

XIANG Zhihua,SHAO Yali   

  1. School of Information Technology,Guangdong Polytechnic College,Zhaoqing 526100, China
  • Received:2017-11-30 Online:2019-11-15 Published:2019-11-15
  • Supported by:
    Science and Technology Project of Guangdong Education Department(201713720010)

摘要:

为解决传统聚类算法无法对高维数据聚类的问题,文中提出了一种结合贪心选择和特征加权的TC-Mean shift高维数据聚类算法。通过对一维数据进行聚类,获得一维数据的聚类结果,再通过加权添加维度聚类,最终获得所有维度数据的聚类,实现对高维数据的聚类。测试结果表明,该算法能够准确地对稀疏的高维数据样本进行聚类,能够处理各种维度的数据,具有良好的实际应用价值。

关键词: 贪心策略, 特征加权, 聚类, 高维数据, Mean shift

Abstract:

In order to solve the problem that traditional clustering algorithms can not cluster high-dimensional data, a high-dimensional data clustering algorithm combining greedy selection and feature weighting was proposed. By clustering one-dimensional feature data, the clustering results of one-dimensional data were obtained first, and then all dimension data were clustered by adding dimension clustering weights to achieve clustering of high-dimensional data. The results showed that the algorithm can accurately cluster sparse high-dimensional data samples and meet the needs of high-dimensional data clustering processing, and had good practical application value.

Key words: greedy strategy, feature weighting, clustering, high-dimensional data, Mean shift

中图分类号: 

  • TP392