DocumentCode
2043680
Title
An investigation into unsupervised clustering techniques
Author
Lee, H.S. ; Younan, N.H.
Author_Institution
Dept. of Electr. & Comput. Eng., Mississippi State Univ., MS, USA
fYear
2000
fDate
2000
Firstpage
124
Lastpage
130
Abstract
The performance of several unsupervised clustering techniques is compared using two clearly separated 3-D data sets that are not separable by any hyperplane. The result shows that the self-organizing feature map can cluster data sets successfully without any prior information of given data while the k-means and the fuzzy k-means algorithm fail to cluster correctly
Keywords
data analysis; data compression; fuzzy systems; matrix algebra; pattern recognition; self-organising feature maps; unsupervised learning; U-matrix method; data compression; exploratory data analysis; fuzzy k-means algorithm; k-means algorithm; self-organizing feature map; separated 3D data sets; simulation; unsupervised clustering techniques; Clustering algorithms; Data analysis; Fuzzy sets; Iterative algorithms; Nearest neighbor searches; Neural networks; Organizing; Partitioning algorithms; Pattern recognition; Signal processing algorithms;
fLanguage
English
Publisher
ieee
Conference_Titel
Southeastcon 2000. Proceedings of the IEEE
Conference_Location
Nashville, TN
Print_ISBN
0-7803-6312-4
Type
conf
DOI
10.1109/SECON.2000.845446
Filename
845446
Link To Document