, Volume 75, Issue 4, pp 359–379 | Cite as

Elastic Principal Graphs and Manifolds and their Practical Applications

  • A. GorbanEmail author
  • A. Zinovyev


Principal manifolds serve as useful tool for many practical applications. These manifolds are defined as lines or surfaces passing through “the middle” of data distribution. We propose an algorithm for fast construction of grid approximations of principal manifolds with given topology. It is based on analogy of principal manifold and elastic membrane. First advantage of this method is a form of the functional to be minimized which becomes quadratic at the step of the vertices position refinement. This makes the algorithm very effective, especially for parallel implementations. Another advantage is that the same algorithmic kernel is applied to construct principal manifolds of different dimensions and topologies. We demonstrate how flexibility of the approach allows numerous adaptive strategies like principal graph constructing, etc. The algorithm is implemented as a C++ package elmap and as a part of stand-alone data visualization tool VidaExpert, available on the web. We describe the approach and provide several examples of its application with speed performance characteristics.

AMS Subject Classifications

62H25 62-07 62-09 68P05 


Principal manifolds elastic functional data analysis data visualization surface modeling 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Aizenberg, L.: Carleman’s formulas in complex analysis: theory and applications. Math. Appl., 244. Kluwer 1993.Google Scholar
  2. Banfield, J. D., Raftery, A. E. 1992Ice flow identification in satellite images using mathematical morphology and clustering about principal curvesJ. Am. Stat. Assoc.87716Google Scholar
  3. Born, M., Huang, K. 1954Dynamical theory of crystal latticesOxford University PressOxfordGoogle Scholar
  4. Cai, W., Shao, X., Maigret, B. 2002Protein-ligand recognition using spherical harmonic molecular surfaces: towards a fast and efficient filter for large virtual throughput screening. J. Mol. Graph. Model2031328CrossRefPubMedGoogle Scholar
  5. Dergachev, V. A., Gorban, A. N., Rossiev, A. A., Karimova, L. M., Kuandykov, E. B., Makarenko, N. G., Steier, P. 2001The filling of gaps in geophysical time series by artificial neural networksRadiocarbon43365371Google Scholar
  6. Dongarra, J., Lumsdaine, A., Pozo, R., Remington, K.: A sparse matrix library in C++ for high performance architectures. In: Proc. 2nd Object Oriented Numerics Conference, pp. 214–218, 1994.Google Scholar
  7. Elmap: C++ package available online: zinovyev/vidaexpert/elmap.Google Scholar
  8. Erwin, E., Obermayer, K., Schulten, K. 1992Self-organizing maps: ordering, convergence properties and energy functionsBiol. Cybern.674755PubMedGoogle Scholar
  9. Gorban, A. N. (ed.): Methods of neuroinformatics (in Russian). Krasnoyarsk State University Press, p. 205, 1998.Google Scholar
  10. Gorban, A. N., Karlin, I. V., Zinovyev, A. Yu.: Invariant grids for reaction kinetics. Physica A 333, 106–154 (2004). Preprint online:–42.html.Google Scholar
  11. Gorban, A. N., Karlin, I. V., Zinovyev, A. Yu.: Constructive methods of invariant manifolds for kinetic problems. Phys. Reports 396(4–6), 197–403 (2004). Preprint online: Scholar
  12. Gorban, A. N., Pitenko, A. A., Zinov’ev, A. Y., Wunsch, D. C. 2001Vizualization of any data using elastic map methodSmart Eng. Syst. Des.11363368Google Scholar
  13. Gorban, A. N., Rossiev, A. A. 1999Neural network iterative method of principal curves for data with gapsJ. Comput. Sys. Sci. Int.38825831Google Scholar
  14. Gorban, A., Rossiev, A., Makarenko, N., Kuandykov, Y., Dergachev, V. 2002Recovering data gaps through neural network methodsInt. J. Geomagnetism Aeronomy3191197Google Scholar
  15. Gorban, A. N., Rossiev, A. A., Wunsch, D. C. II: Neural network modeling of data with gaps: Method of principal curves, Carleman’s formula, and other. In: USA–NIS Neurocomputing Opportunities Workshop, Washington, July 1999 (Associated with IJCNN’99). Preprint online: Scholar
  16. Gorban, A. N., Zinovyev, A. Yu.: Visualization of data by method of elastic maps and its applications in genomics, economics and sociology. Preprint of Institut des Hautes Etudes Scientiques, M/01/36, 2001. Scholar
  17. Gorban, A. N., Zinovyev, A. Yu. 2001Method of elastic maps and its applications in data visualization and data modelingInt. J. Comput. Anticipatory Syst. CHAOS12353369Google Scholar
  18. Gorban, A. N., Zinovyev, A. Yu., Pitenko, A. A. 2000Visualization of data using method of elastic maps (in Russian)Informatsionnie Technologii62635Google Scholar
  19. Gorban, A. N., Zinovyev, A. Yu., Pitenko, A. A. 2002Visualization of data. Method of elastic map (in Russian)Neurocomputers41930Google Scholar
  20. Gorban, A. N., Zinovyev, A. Yu., Wunsch, D. C.: Application of the method of elastic maps in analysis of genetic texts. In: Proc. Int. Joint Conf. on Neural Networks (IJCNN), Portland, July 20–24, 2003.Google Scholar
  21. Gusev, A. 2004Finite element mapping for spring network representations of the mechanics of solidsPhys. Rev. Lett.93034302CrossRefPubMedGoogle Scholar
  22. Hastie, T.: Principal curves and surfaces. PhD Thesis, Stanford University, 1984.Google Scholar
  23. Hastie, T., Stuetzle, W. 1989Principal curvesJ. Am. Stat. Assoc.84502516Google Scholar
  24. Kohonen, T. 1982Self-organized formation of topologically correct feature mapsBiol. Cybern.435969CrossRefGoogle Scholar
  25. Kégl, B.: Principal curves: learning, design, and applications. PhD Thesis, Concordia University, Canada, 1999.Google Scholar
  26. Kégl, B., Krzyzak, A. 2002Piecewise linear skeletonization using principal curvesIEEE Trans. Pattern Anal. Machine Intell.245974CrossRefGoogle Scholar
  27. Kégl, B., Krzyzak, A., Linder, T., Zeger, K.: A polygonal line algorithm for constructing principal curves. Neural Inf. Processing Sys. pp. 501–507 (1999).Google Scholar
  28. Kégl, B., Krzyzak, A., Linder, T., Zeger, K. 2000Learning and design of principal curvesIEEE Trans. Pattern Anal. Machine Intell.22281297CrossRefGoogle Scholar
  29. LeBlanc, M., Tibshirani, R. 1994Adaptive principal surfacesJ. Amer. Stat. Assoc.895364Google Scholar
  30. Mandal, C., Qin, H., Vemuri, B. C. 2000A novel FEM-based dynamic framework for subdivision surfacesComp. Aided Des.32479497CrossRefGoogle Scholar
  31. Mulier, F., Cherkassky, V. 1995Self-organization as an iterative kernel smoothing processNeural Comput.711651177PubMedGoogle Scholar
  32. Ritter, H., Martinetz, T., Schulten, K.: Neural computation and self-organizing maps: an introduction. Addison-Wesley Reading, Massa. 1992.Google Scholar
  33. Roweis, S., Saul, L. K. 2000Nonlinear dimensionality reduction by locally linear embeddingScience29023232326CrossRefPubMedGoogle Scholar
  34. Sayle, R., Bissell, A.: RasMol: a program for fast realistic rendering of molecular structures with shadows. In: Proc. 10th Eurographics UK’92 Conf., University of Edinburgh, Scotland, 1992.Google Scholar
  35. Stanford, D., Raftery, A. E. 2000Principal curve clustering with noiseIEEE Trans. Pattern Anal. Machine Intell.22601609CrossRefGoogle Scholar
  36. Tenenbaum, J. B., DeSilva, V., Langford, J. C. 2000A global geometric framework for nonlinear dimensionality reductionScience29023192323CrossRefPubMedGoogle Scholar
  37. Van Gelder, A., Wilhelms, J.: Simulation of elastic membranes and soft tissue with triangulated spring meshes. Technical Report: UCSC-CRL-97–12, 1997.Google Scholar
  38. Verbeek, J. J., Vlassis, N., Krose, B.: A k-segments algorithm for finding principal curves. Technical report, 2000. Online: Scholar
  39. VidaExpert: Stand-alone application for multidimensional data visualization. Available online: zinovyev/vidaexpert/vidaexpert.htm.Google Scholar
  40. Xie, H., Qin, H.: A physics-based framework for subdivision surface design with automatic rules control. In: Proc. 10th Pacific Conf. on Computer Graphics and Applications (Pacific Graphics 2002), IEEE Press, 304–315, 2002.Google Scholar
  41. Zinovyev, A.: Visualization of multidimensional data. Krasnoyarsk State University Press Publ., 2000.Google Scholar
  42. Zinovyev, A. Yu., Gorban, A. N., Popova, T. G. 2003Self-organizing approach for automated gene identificationOpen Sys. Inf. Dyn.10321333CrossRefMathSciNetGoogle Scholar
  43. Zinovyev, A. Yu., Pitenko, A. A., Popova, T. G. 2002Practical applications of the method of elastic maps (in Russian)Neurocomputers43139Google Scholar

Copyright information

© Springer-Verlag Wien 2005

Authors and Affiliations

  1. 1.University of LeicesterLeicesterUK
  2. 2.Institut CurieParisFrance

Personalised recommendations