Neural Processing Letters

, Volume 5, Issue 3, pp 193–199 | Cite as

Self-Organizing Maps for the Analysis of Complex Movement Patterns

  • H.U. Bauer
  • W. Schöllhorn
Article

Abstract

We apply the Self-Organizing-Map-algorithm (SOM) as a central processing step in a new scheme for the characterisation of movement patterns of athletes. Due to its non-linear dimension reduction capabilities, the SOM outperforms a direct processing of the data as well as preprocessing using principal component analysis. Our results open the way to an objective assessment of movement patterns, with possible applications in the sport sciences as well as in medicine.

cluster analysis dimension-reduction movement pattern neighborhood preservation SOM 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    S.H. Holzreiter, M.E. Köhle, “Assessment of gait patterns using neural networks”, J. Biomech. Vol. 26, pp. 645–651, 1993.Google Scholar
  2. 2.
    T. Kohonen, “Self-Organization Maps”, Springer Series in Information Science 30, Springer: Berlin, Heidelberg, New York, 1995.Google Scholar
  3. 3.
    H. Ritter, T. Martinetz, K. Schulten, “Neural Computation and Self-Organizing Maps”, Addison-Wesley: Reading, Mass., 1992.Google Scholar
  4. 4.
    W. Schöllhorn, “Orthogonale Referenz-Funktionen zur Erkennung von Bewegungsmustern”, in J. Krug, Zeitreihenanalyse - multiple statistische Verfahren in der Trainingswissenschaft, Köln, Sport Buch Strauss, pp. 103–120, 1996.Google Scholar
  5. 5.
    M.J. Norusius/ SPSS Inc., SPSS/PC+4.0 Statistics, Chicago, IL 60611.Google Scholar
  6. 6.
    H. Haken, “Principles of Brain Functioning”, Springer, Berlin, 1996.Google Scholar
  7. 7.
    H.-U. Bauer, K. Pawelzik, “Quantifying the neighbourhood preservation of Self-Organizing Feature Maps”, IEEE Trans. on Neural Netw., Vol. 3, No. 4, pp. 570–579, 1992.Google Scholar
  8. 8.
    J.C. Bezdek, N.R. Pal, “An index of topology preservation for feature extraction”, Patt. Rec. Vol. 28, pp.381–391, 1995.Google Scholar
  9. 9.
    Th. Villmann, R. Der, Th. Martinetz, “A novel approach to measure the topology preservation of feature maps”, in M. Marinaro, P.G. Morasso, (eds) Proc. of the International Conference on Artificial Neural Networks 1994, Sorrento, Italy, Springer Verlag, London, pp. 298–301, 1994.Google Scholar
  10. 10.
    H.-U. Bauer, M. Herrmann, Th. Villmann, “Topology preservation in neural maps”, submitted to IEEE Trans. Sign. Proc., 1996.Google Scholar
  11. 11.
    W. Schöllhorn, H.-U. Bauer, in preparation, 1996.Google Scholar

Copyright information

© Kluwer Academic Publishers 1997

Authors and Affiliations

  • H.U. Bauer
    • 1
    • 2
  • W. Schöllhorn
    • 3
  1. 1.Max-Planck-Institut für StrömungsforschungGöttingenGermany
  2. 2.SFB 185 ‘Nichtllineare Dynamik’JWG-UniversitätFrankfurtGermany
  3. 3.Institut für BiomechanikJWG-UniversitätFrankfurtGermany

Personalised recommendations