Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Journal of Changchun University》 2017-08
Add to Favorite Get Latest Update

Design of Center Random Drift( CRD) KMeans Clustering Algorithm

YIN Xiaoli;Department of Basic Teaching,Business College of Shanxi University;  
KMeans clustering algorithm will probably generate isolated points during clustering process,once this happens,it will seriously deteriorate the effect of the clustering algorithm. In order to avoid generating isolated points,an algorithm called center random drift( CRD) KMeans clustering is designed,which will firstly determine whether the cluster point is isolated,and if it is,it will be replaced by using CRD algorithm,so as to avoid the isolated point effectively. Through comparative experiments of image clustering in Matlab environment,it is found that there is no significant difference between the new algorithm and the traditional KMeans algorithm for the colorful pictures,while for a picture with a single color,the traditional KMeans clustering algorithm's clustering effect is poor,but the new algorithm can still work effectively.
【Fund】: 国家社会科学规划基金(16BTJ034)
【CateGory Index】: TP311.13
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved