Classifiers in almost empty spaces

Author

Duin, Robert P W

Author_Institution

Dept. of Appl. Phys., Delft Univ. of Technol., Netherlands

Volume

2

fYear

2000

fDate

2000

Firstpage

1

Abstract

Recent developments in defining and training statistical classifiers make it possible to build reliable classifiers in very small sample size problems. Using these techniques advanced problems may be tackled, such as pixel based image recognition and dissimilarity based object classification. It can be explained and illustrated how recognition systems based on support vector machines and subspace classifiers circumvent the curse of dimensionality, and even may find nonlinear decision boundaries for small training sets represented in Hilbert space

Keywords

Hilbert spaces; decision theory; learning automata; object recognition; pattern classification; statistical analysis; Hilbert space; decision theory; dimensionality; dissimilarity; image recognition; kernel mapping; object recognition; pattern classification; statistical classifiers; support vector machines; training sets; Hilbert space; Image databases; Image recognition; NIST; Pattern recognition; Physics; Space technology; Support vector machine classification; Support vector machines; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Pattern Recognition, 2000. Proceedings. 15th International Conference on

Conference_Location

Barcelona

ISSN

1051-4651

Print_ISBN

0-7695-0750-6

Type

conf

DOI

10.1109/ICPR.2000.906006

Filename

906006