首页 | 本学科首页   官方微博 | 高级检索  
   检索      

基于网格的聚类挖掘算法
引用本文:郑世明,苗壮,赵波,赵志宁.基于网格的聚类挖掘算法[J].军械工程学院学报,2011(3):65-68.
作者姓名:郑世明  苗壮  赵波  赵志宁
作者单位:[1]解放军理工大学指挥自动化学院,江苏南京210007 [2]军械工程学院光学与电子工程系,河北石家庄050003 [3]炮兵指挥学院基础部,河北张家口075100
摘    要:为了提高海量数据挖掘效率,研究了一种基于网格环境下的分布式聚类(Prejudge-Based Distributed Clus-tering,PBDC)算法,并引入距离、模和内积的概念,在聚类之前进行预判断,减少了不必要的计算开销。在此基础上提出了一种分布式并行化聚类(Distributed Parallel Clustering,DPC)算法,将其嵌入到Weka4ws中,以开源数据挖掘类库Weka为底层支持环境,构建网格环境下的分布式数据挖掘体系,同时进行仿真实验。实验结果表明:该算法对于网格环境下海量数据的分布式聚类具有良好的效果。

关 键 词:网格  分布式  聚类  数据挖掘  Weka4ws

Clustering Mining Algorithm Based on Grid
ZHENG Shi-ming,MIAO Zhuang,ZHAO Bo,ZHAO Zhi-ning.Clustering Mining Algorithm Based on Grid[J].Journal of Ordnance Engineering College,2011(3):65-68.
Authors:ZHENG Shi-ming  MIAO Zhuang  ZHAO Bo  ZHAO Zhi-ning
Institution:1.Institute of Command Automation,PLA University of Science and Technology,Nanjing 210007,China;2.Department of Optics and Electronics Engineering,Ordnance Engineering College,Shijiazhuang 050003,China;3.Department of Basic Courses,Artillery Command Academy,Zhangjiakou 075100,China)
Abstract:In order to improve clustering efficiency,this paper suggests a distributed clustering algorithm PBDC under grid environment by introducing the fundamental conception of distance,mode,and inner product.Making a pre-judgement before clustering reduces unnecessary computation,and on the basis of this we present a new distributed parallel clustering algorithm DPC by dint of Weka Library and design a distributed data mining architecture in grid environment.Finally we validates the algorithm with the distributed clustering based on Weka4ws.
Keywords:grid  distributed  clustering  data mining  Weka4ws
本文献已被 维普 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