Title :
Automatic pattern recognition by similarity representations
Author :
Pekalska, E. ; Duin, R.P.W.
Author_Institution :
Dept. of Appl. Phys., Delft Univ. of Technol., Netherlands
fDate :
2/1/2001 12:00:00 AM
Abstract :
The automatic recognition of objects may benefit from using a similarity representation instead of the traditional approach based on features. It is shown that the common use of nearest neighbour classifiers for similarity representations may be improved significantly by other classification rules with respect to recognition accuracy as well as computational complexity
Keywords :
computational complexity; pattern classification; automatic pattern recognition; classification rules; computational complexity; nearest neighbour classifiers; recognition accuracy; similarity representations;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:20010121
