Subclass problem-dependent design for error-correcting output codes

Sergio Escalera, David M.J. Tax, Oriol Pujol, Petia Radeva, Robert P.W. Duin

    Research output: Contribution to journalArticleResearchpeer-review

    97 Citations (Scopus)

    Abstract

    A common way to model multi-class classification problems is by means of Error-Correcting Output Codes (ECOC). Given a multi-class problem, the ECOC technique designs a codeword for each class, where each position of the code identifies the membership of the class for a given binary problem. A classification decision is obtained by assigning the label of the class with the closest code. One of the main requirements of the ECOC design is that the base classifier is capable of splitting each sub-group of classes from each binary problem. However, we can not guarantee that a linear classifier model convex regions. Furthermore, non-linear classifiers also fail to manage some type of surfaces. In this paper, we present a novel strategy to model multi-class classification problems using sub-class information in the ECOC framework. Complex problems are solved by splitting the original set of classes into sub-classes, and embedding the binary problems in a problem-dependent ECOC design. Experimental results show that the proposed splitting procedure yields a better performance when the class overlap or the distribution of the training objects conceil the decision boundaries for the base classifier. The results are even more significant when one has a sufficiently large training size. © 2008 IEEE.
    Original languageEnglish
    Pages (from-to)1041-1054
    JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
    Volume30
    Issue number6
    DOIs
    Publication statusPublished - 1 Jun 2008

    Keywords

    • Embedding of dichotomizers
    • Error-correcting output codes
    • Multiclass classification
    • Subclasses

    Fingerprint Dive into the research topics of 'Subclass problem-dependent design for error-correcting output codes'. Together they form a unique fingerprint.

    Cite this