Possibilities of zero-error recognition by dissimilarity representations

Conference Paper (2002)
Author(s)

Bob Duin (TU Delft - ImPhys/Quantitative Imaging)

EM Pekalska (TU Delft - ImPhys/Quantitative Imaging)

Research Group
ImPhys/Quantitative Imaging
More Info
expand_more
Publication Year
2002
Research Group
ImPhys/Quantitative Imaging
Pages (from-to)
20-32
ISBN (print)
972-98050-8-3

Abstract

Feature based approaches to pattern recognition suffer from the fact that feature representations of different classes of objects may overlap. This is the consequence of reducing the description of an object to a feature vector. As a result an error free recognition system is even asymptotically (for infinite training sizes) impossible. In this paper it is argued that this limitation does not hold for dissimilarity based representations. Suggestions are made how this may be exploited in practice.

No files available

Metadata only record. There are no files for this record.