Typhoon Analysis and Data Mining with Kernel Methods

  • Asanobu Kitamoto
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2388)


The analysis of the typhoon is based on the manual pattern recognition of cloud patterns on meteorological satellite images by human experts, but this process may be unstable and unreliable, and we think could be improved by taking advantage of both the large collection of past observations and the state-of-the-art machine learning methods, among which kernel methods, such as support vector machines (SVM) and kernel PCA, are the focus of the paper. To apply the ”learning-from-data” paradigm to typhoon analysis, we built the collection of more than 34,000 well-framed typhoon images to be used for spatio-temporal data mining of typhoon cloud patterns with the aim of discovering hidden and unknown regularities contained in large image databases. In this paper, we deal with the problem of visualizing and classifying typhoon cloud patterns using kernel methods. We compare preliminary results with baseline algorithms, such as principal component analysis and a k-NN classifier, and discuss experimental results with the future direction of research.


Support Vector Machine Central Pressure Kernel Method Human Expert Cloud Fraction 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    V. N. Vapnik. Statistical Learning Theory. John Wiley & Sons, Inc., 1998.Google Scholar
  2. 2.
    B. Schölkopf, C. J. C. Burges, and A. J. Smola, editors. Advances in Kernel Methods. The MIT Press, 1999.Google Scholar
  3. 3.
    B. Schölkopf and A. J. Smola. Learning with Kernels. The MIT Press, 2002.Google Scholar
  4. 4.
    A. Kitamoto. Spatio-temporal data mining for typhoon image collection. Journal of Intelligent Information Systems, 19(1), 2002. (in press).Google Scholar
  5. 5.
    A. Kitamoto. IMET: Image mining environment for typhoon analysis and prediction. In C. Djeraba, editor, Multimedia Data Mining. Kluwer Academic Publishers, 2002. (in press).Google Scholar
  6. 6.
    V. F. Dvorak. Tropical cyclone intensity analysis using satellite data. NOAA Technical Report NESDIS, 11:1–47, 1984.Google Scholar
  7. 7.
    E. N. Lorenz. The Essence of Chaos. University of Washington Press, 1993.Google Scholar
  8. 8.
    T. Hastie, R. Tibshirani, and J. Friedman. The Elements of Statistical Learning. Springer, 2001.Google Scholar
  9. 9.
    A. Kitamoto. The development of typhoon image database with content-based search. In Proceedings of the 1st International Symposium on Advanced Informatics, pages 163–170, 2000.Google Scholar
  10. 10.
    R. S. T. Lee and J. N. K. Liu. An automatic satellite interpretation of tropical cyclone patterns using elastic graph dynamic link model. Journal of Pattern Recognition and Artificial Intelligence, 13(8):1251–1270, 1999.CrossRefGoogle Scholar
  11. 11.
    L. Zhou, C. Kambhamettu, and D. B. Goldgof. Extracting nonrigid motion and 3D structure of hurricanes from satellite image sequences without correspondences. In Proc. of Conference on Computer Vision and Pattern Recognition. IEEE, 1999.Google Scholar
  12. 12.
    D. S. Wilks. Statistical Methods in the Atmospheric Sciences. Academic Press, 1995.Google Scholar
  13. 13.
    E. Anderson, Z. Bai, C. Bischof, S. Blackford, J. Demmel, J. Dongarra, J. Du Croz, A. Greenbaum, S. Hammarling, A. McKenney, and D. Sorensen. LAPACK Users’ Guide. Society for Industrial and Applied Mathematics, Philadelphia, PA, third edition, 1999.Google Scholar
  14. 14.
    C. W. Hsu and C. J. Lin. A simple decomposition method for support vector machines. Machine Learning, 46:291–314, 2002.zbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Asanobu Kitamoto
    • 1
  1. 1.National Institute of Informatics (NII)TokyoJapan

Personalised recommendations