首页 | 本学科首页   官方微博 | 高级检索  
     

基于集群环境的K-Means聚类算法的并行化
引用本文:王辉,张望,范明. 基于集群环境的K-Means聚类算法的并行化[J]. 河南科技大学学报(自然科学版), 2008, 29(4)
作者姓名:王辉  张望  范明
作者单位:河南科技大学电子信息工程学院,河南,洛阳,471003;河南科技大学现代教育技术与信息中心,河南,洛阳,471003
基金项目:国家自然科学基金,教育部科学技术基金
摘    要:
K-Means聚类算法在面对海量数据时,时间和空间的复杂性已成为K-Means聚类算法的瓶颈.在充分研究传统K-Means聚类算法的基础上,提出了基于集群环境的并行K-Means聚类算法的设计思想,给出了其加速比估算公式,并通过实验证明了该算法的正确性和有效性.

关 键 词:集群  并行  K-Means聚类算法

Research on Parallelism of K-Means Clustering Algorithm Based on Cluster
WANG Hui,ZHANG Wang,FAN Ming. Research on Parallelism of K-Means Clustering Algorithm Based on Cluster[J]. Journal of Henan University of Science & Technology:Natural Science, 2008, 29(4)
Authors:WANG Hui  ZHANG Wang  FAN Ming
Affiliation:WANG Huia,ZHANG Wanga,FAN Mingb(a.Electronic Information Engineering College,b.Modern Education Technology & Information Center,Henan University of Science & Technology,Luoyang 471003,China)
Abstract:
The complexity of time and spatial is becoming the difficulty of K-Means clustering algorithm while it deals with the huge amounts of data sets.Based on the study of the traditional K-Means clustering algorithm,the design concept of the parallel K-Means algorithm is discussed and a formula of the speedup ratio is proposed.The accuracy and validity of the algorithm through experiments are proved.
Keywords:Cluster  Parallelism  K-Means clustering algorithm  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号