Title :
A New Method for Non-spherical and Multi-density Clustering
Author :
Yang, Xiaobing ; He, Lingmin ; Lu, Huijuan
Author_Institution :
Comput. Sci. Dept., China Jiliang Univ., Hangzhou, China
Abstract :
Cluster analysis based on Delaunay diagrams has been widely researched in all kinds of fields. In this article, a new method is proposed to treat with datasets of non-spherical and multi-density clustering. As well as, the problem of noise data is solved by defining the conceptions of In_Points and Out_Points in this method. The above points was illustrated by two simulation experiments.
Keywords :
mesh generation; noise; pattern clustering; Delaunay diagrams; In_Points conception; Out_Points conception; multidensity clustering; noise data problem; nonspherical clustering; Algorithm design and analysis; Application software; Clustering algorithms; Clustering methods; Computer science; Helium; Information analysis; Information technology; Multi-stage noise shaping; Shape; Delaunay Diagrams; IPOP algorithm; cluster analysis; data mining;
Conference_Titel :
Intelligent Information Technology Application, 2009. IITA 2009. Third International Symposium on
Conference_Location :
Nanchang
Print_ISBN :
978-0-7695-3859-4
DOI :
10.1109/IITA.2009.376
