Title :
Unsupervised pattern recognition methods for interval data using non-quadratic distances
Author :
de Carvalho, F.A.T. ; de Souza, R.M.C.R.
Author_Institution :
Center of Informatics, Univ. Fed. de Pernambuco, Recife, Brazil
fDate :
3/6/2003 12:00:00 AM
Abstract :
Unsupervised pattern recognition methods for interval data using a dynamic cluster algorithm are presented. Two methods are considered: one with adaptive distances and the other without. They can be applied to image segmentation. The clustering outputs are compared using an external index. The best results are furnished by adaptive methods.
Keywords :
adaptive signal processing; image segmentation; pattern clustering; adaptive distances; adaptive methods; clustering outputs comparison; dynamic cluster algorithm; external index; image segmentation; interval data; unsupervised pattern recognition methods;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:20030304
