In these years, we often deal with an enormous amount of data or a data stream in a large variety of pattern recognition tasks. As a promising approach for economising the amount, we have previously defined a volume prototype as a geometric configuration that represents some data points inside and proposed a single-pass algorithm for finding them. In this paper, we analyze the convergence behavior of volume prototypes in high-dimensional cases. In addition, we show the applicability of volume prototypes to high-dimensional classification problems.


Covariance Matrix Mixture Model Data Stream Recognition Rate Behavior Analysis 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Tabata, K., Kudo, M.: Information compression by volume prototypes. The IEICE Technical Report, PRMU 106(429), 25–30 (2006) (in Japanese) Google Scholar
  2. 2.
    Kaymak, U., Setnes, M.: Fuzzy clustering with volume prototypes and adaptive cluster merging. IEEE Transactions on Fuzzy Systems 10(6), 705–712 (2002)CrossRefGoogle Scholar
  3. 3.
    Fukunaga, K.: Introduction to Statistical Pattern Recognition. Academic Press, London (1990)zbMATHGoogle Scholar
  4. 4.
    Asuncion, A., Newman, D.: UCI machine learning repository (2007)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Maiko Sato
    • 1
  • Mineichi Kudo
    • 1
  • Jun Toyama
    • 1
  1. 1.Division of Computer Science Graduate School of Information Science and TechnologyHokkaido UniversitySapporoJapan

Personalised recommendations