Prototype Selection for Finding Efficient Representations of Dissimilarity Data

EM Pekalska, RPW Duin

    Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

    9 Citations (Scopus)

    Abstract

    The nearest neighbor (NN) rule is a simple and intuitive method for solving classification problems. Originally, it uses distances to the complete training set. It performs well, however, it is sensitive to noisy objects, due to its operation on local neighborhoods only. A more global approach is possible by mapping the distance data onto a pseudo- Euclidean space, such that the distances are preserved as well as possible. Then, a classifier built in such a space can outperform the NN rule. However, again all objects from the training set are used for a projection of new data. This paper addresses the issue of reducing the training set while possibly preserving the original structure of the mapped data. Some criteria are introduced and evaluated against two problems, polygon recognition and digit recognition. Our experiments show that the representation mismatch criterion is beneficial for the applications considered. Moreover, the linear classifier built in the pseudo- Euclidean space, determined by 20%¿25% of the training objects, outperforms the NN rule based on all of them.
    Original languageUndefined/Unknown
    Title of host publicationICPR16, Proceedings
    EditorsR Kasturi, D Laurendeau, C Suen
    Place of PublicationLos Alamitos, CA
    PublisherIEEE
    Pages37-40
    Number of pages4
    ISBN (Print)0-7695-1696-3
    Publication statusPublished - 2002
    Event16th International Conference on Pattern Recognition (Quebec City, Canada), vol. III - Los Alamitos, CA
    Duration: 11 Aug 200215 Aug 2002

    Publication series

    Name
    PublisherIEEE Computer Society Press
    NameInternational Conference on Pattern Recognition
    Volume3
    ISSN (Print)1051-4651

    Conference

    Conference16th International Conference on Pattern Recognition (Quebec City, Canada), vol. III
    Period11/08/0215/08/02

    Bibliographical note

    ISSN 1051-4651, phpub 45

    Keywords

    • conference contrib. refereed
    • Conf.proc. > 3 pag

    Cite this