Novel Metrics Formulate with Well-Initialized Setting

  • Xiangyang You
  • Xiangsheng Rong
  • Fujiang Huo
  • Ming Xu
  • Yuanzheng Zhang
Conference paper
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 205)

Abstract

A new method rooted from Vector Quantization (VQ) is proposed in this paper. The algorithm is encoded with a well-defined initialized setting of prototypes. And the algorithm is involved by employment of prototype-customized Metrics, so the algorithm is named as WDVQ. WDVQ looks for initialized points of significant meaning from dataset and derives the first population of prototypes based on them. It creates 1-vs-rest SVMs model and uses its decision functions to learn individual metric for each prototype. The prototype updating takes two fashions rather than a single manner that is used in classical VQ. Experiments are conducted on real datasets to check the quality of initial prototypes, customized metrics as well as WDVQ and show the fine performance of the proposed strategies and algorithm.

Keywords

Offline and online membership function SVM Weighted schema Parameter tuning 

References

  1. 1.
    Kohonen T (1995) Learning vector quantization, In: The handbook of brain theory and neural networks, vol 12, MIT Press, Cambridge, pp: 537–540 Google Scholar
  2. 2.
    Bezdek JC (1981) Pattern recognition with fuzzy objective function algorithms. Plenum, New York vol 20 pp: 48–53Google Scholar
  3. 3.
    Hathaway RJ, Bezdek JC (1995) Optimization of clustering criteria by reformulation, In: IEEE trans. fuzzy system, IEEE computational intelligence society vol 3 pp: 241–246Google Scholar
  4. 4.
    Bradley PS, Fayyad UM (1998) Refining initial points for K-means clustering. In: Proceedings of 15th international conference on machine learning 23:91–99Google Scholar
  5. 5.
    Karayiannis NB (2000) Generalized fuzzy C-means algorithms. J Intell Fuzzy Syst vol 8(1) pp: 63–81 Google Scholar
  6. 6.
    Karayiannis NB, Intell J (1997) Fuzzy partition entropies and entropy constrained fuzzy clustering algorithms. Fuzzy Syst vol 5(2) pp: 103–111Google Scholar
  7. 7.
    Kohonen T (2001) Self organization maps. Springer, New YorkGoogle Scholar
  8. 8.
    Gath I, Geva AB (1989) Unsupervised optimal fuzzy clustering. IEEE Trans Pattern Anal Mach Intell vol 11 pp: 773–781Google Scholar

Copyright information

© Springer-Verlag London 2013

Authors and Affiliations

  • Xiangyang You
    • 1
  • Xiangsheng Rong
    • 1
  • Fujiang Huo
    • 2
  • Ming Xu
    • 2
  • Yuanzheng Zhang
    • 1
  1. 1.Training DepartmentXuzhou Air Force College of P. L. AXuzhouChina
  2. 2.Department of Logistic CommandXuzhou Air Force College of P. L. AXuzhouChina

Personalised recommendations