Determining Relevant Input Dimensions for the Self Organizing Map

  • Thorsten Bojer
  • Barbara Hammer
  • Marc Strickert
  • Thomas Villmann
Part of the Advances in Soft Computing book series (AINSC, volume 19)


We propose a method to determine the relevance of the different input dimensions for a self organizing map (SOM). First, a growing self organizing map is adapted to the data. Afterwards, the effect of the input dimensions on the clustering or the topology of the SOM, respectively, is computed and the data dimensions which are ranked low are pruned. The algorithm is applied to real life satellite image data. The results are verified via visualizing the data in RGB-images as well as explicitely computing the classification error.


Receptive Field Input Dimension Pruning Method Topology Preservation Codebook Vector 
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.
    H.-U. Bauer and K. R. Pawelzik. Quantifying the neighborhood preservation of Self-Organizing Feature Maps. IEEE Trans. on Neural Networks, 3 (4): 570–579, 1992.CrossRefGoogle Scholar
  2. 2.
    H.-U. Bauer and T. Villmann. Growing a Hypercubical Output Space in a Self-Organizing Feature Map. IEEE Transactions on Neural Networks, 8 (2): 218–226, 1997.CrossRefGoogle Scholar
  3. 3.
    B. Fritzke. Growing grid: a self-organizing network wirh constant neighborhood range and adaptation strength. Neural Processing Letters, 2 (5): 9–13, 1995.CrossRefGoogle Scholar
  4. 4.
    P. Grassberger and I. Procaccia. Measuring the strangeness of strange attractors. Physica, 9D: 189–208, 1983.MathSciNetMATHGoogle Scholar
  5. 5.
    B. Hammer and T. Villmann. Input pruning for neural gas architectures. In: M.Verleysen, editro, ESANN’2001, pages 271–276, 2001.Google Scholar
  6. 6.
    T. Honkela, S. Kaski, T. Kohonen, and K. Lagus. Self-organizing maps of very large document collections: Justification for the WEBSOM method. In I. Balderjahn, R. Mathar, and M. Schader, editors, Classification, Data Analysis, and Data Highways, pages 245252. Springer, Berlin, 1998.Google Scholar
  7. 7.
    S. Kaski. Dimensionality reduction by random mapping: fast similarity computation for clustering. In Proceedings of IJCNN’92, pages 413-. 418, 1998.Google Scholar
  8. 8.
    T. Kohonen. Self-Organizing Maps. Springer, 1997.Google Scholar
  9. 9.
    T. Martinetz and K. Schulten. Topology representing networks. Neural Networks, 7 (3): 507–522, 1993.CrossRefGoogle Scholar
  10. 10.
    U. Matecki. Automatische Merkmalsauswahl fiir Neuronale Netze mit Anwendung in der pixelbezogenen Klassifikation von Bildern. Shaker, 1999.Google Scholar
  11. 11.
    A. Meyering and H. Ritter. Learning 3D-shape-perception with local linear maps. In Proceedings of IJCNN’92, pages 432–436, 1992.Google Scholar
  12. 12.
    R. Nock and M. Sebban. Sharper bounds for the hardness of prototype and feature selection. In H. Arimura, S. Jain, and A. Sharma, editors, Algorihmic Learning Theory, pages 224–237. Springer, 2000.Google Scholar
  13. 13.
    A. Ultsch. Self-organizing neural networks for visualization and classification. In O. Opitz, B. Lausen, and R. Klar, editors, Information and Classification, pages 307313, London, UK, 1993. Springer.Google Scholar
  14. 14.
    T. Villmann, R. Der, M. Herrmann, and T. Martinetz. Topology Preservation in Self-Organizing Feature Maps: Exact Definition and Measurement. IEEE Transactions on Neural Networks, 8 (2): 256–266, 1997.CrossRefGoogle Scholar
  15. 15.
    T. Villmann and E. Merenyi. Extensions and modifications of SOM and its application in satellite remote sensoring processing. In H. Bothe and R.Rojas, editors, Neural Computation 2000, pages 765–771, Zürich, 2000. ICSC Academic Press.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Thorsten Bojer
    • 1
  • Barbara Hammer
    • 1
  • Marc Strickert
    • 1
  • Thomas Villmann
    • 2
  1. 1.Department of Mathematics/Computer ScienceUniversity of OsnabrückOsnabrückGermany
  2. 2.Clinic for Psychotherapy and Psychosomatic MedicineUniversity of LeipzigLeipzigGermany

Personalised recommendations