A New Clustering Method Based on References, Density and Neural Network

Author

Gou, Hong-tu ; Xu, Jian-Suo ; Wang, Li

Author_Institution

Sch. of Manage., Tianjin Univ.

fYear

2006

fDate

13-16 Aug. 2006

Firstpage

1060

Lastpage

1065

Abstract

In this paper, a new clustering algorithm that is called RDVS (clustering using references and density by ViSOM) is presented to overcome the shortcomings of clustering methods based on density or neural network. The creativity of RDVS is capturing the shape and extent of a cluster by references and their densities, and then analyzes them by ViSOM. RDVS keeps the ability of density-based clustering method´s good features and it can give a visual clustering results. Both theory analysis and experimental results confirm that RDVS can discover clusters with arbitrary shape and is insensitive to noise data, and its executing efficiency is much higher than ViSOM

Keywords

data mining; data visualisation; pattern clustering; self-organising feature maps; RDVS clustering algorithm; ViSOM neural network; cluster discovery; density-based clustering method; visualization self-organizing map; Clustering algorithms; Clustering methods; Computer network management; Computer networks; Conference management; Cybernetics; Data mining; Machine learning; Machine learning algorithms; Neural networks; Shape; Technology management; Clustering; data mining; density; neural network; reference;

fLanguage

English

Publisher

ieee

Conference_Titel

Machine Learning and Cybernetics, 2006 International Conference on

Conference_Location

Dalian, China

Print_ISBN

1-4244-0061-9

Type

conf

DOI

10.1109/ICMLC.2006.258560

Filename

4028220