Spectral Clustering Based on k-Nearest Neighbor Graph

  • Małgorzata Lucińska
  • Sławomir T. Wierzchoń
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7564)


Finding clusters in data is a challenging task when the clusters differ widely in shapes, sizes, and densities. We present a novel spectral algorithm Speclus with a similarity measure based on modified mutual nearest neighbor graph. The resulting affinity matrix reflex the true structure of data. Its eigenvectors, that do not change their sign, are used for clustering data. The algorithm requires only one parameter – a number of nearest neighbors, which can be quite easily established. Its performance on both artificial and real data sets is competitive to other solutions.


Spectral clustering nearest neighbor graph signless Laplacian 


  1. 1.
    Chen, Y., Jensen, C.D., Gray, E., Seigneur, J.M.: Risk Probability Estimating Based on Clustering, Technical Report No. TCD-CS-2003-17, Trinity College Dublin (2003)Google Scholar
  2. 2.
    Cvetković, D.: Signless Laplacians and line graphs. Bull. Acad. Serbe Sci. Arts, Cl. Sci. Math. Natur., Sci. Math. 131(30), 85–92 (2005)zbMATHGoogle Scholar
  3. 3.
    Deepak, V., Meila, M.: Comparison of Spectral Clustering Methods. UW TR CSE-03-05-01 (2003)Google Scholar
  4. 4.
    Elon, Y.: Eigenvectors of the discrete Laplacian on regular graphs a statistical approach. J. Phys. A: Math. Theor. 41 (2008)Google Scholar
  5. 5.
    Fischer, I., Poland, J.: Amplifying the Block Matrix Structure for Spectral Clustering. Technical Report No. IDSIA-03-05, Telecommunications Lab (2005)Google Scholar
  6. 6.
    Jain, A., Murty, M., Flynn, P.: Data clustering: A review. ACM Computing Surveys 31, 264–323 (1999)CrossRefGoogle Scholar
  7. 7.
    Jain, A.: Data clustering: 50 years beyond K-means. Pattern Recognition Letters 31, 651–666 (2010)CrossRefGoogle Scholar
  8. 8.
    MacQueen, L.: Some methods for classification and analysis of multivariate observations. In: LeCam, L., Neyman, J. (eds.) 5th Berkeley Symposium on Mathematical Statistics and Probabilitz, vol. 1, pp. 281–297. University of California Press, Berkeley (1967)Google Scholar
  9. 9.
    Maier, M., Hein, M., von Luxburg, U.: Cluster Identification in Nearest-Neighbor Graphs. In: Hutter, M., Servedio, R.A., Takimoto, E. (eds.) ALT 2007. LNCS (LNAI), vol. 4754, pp. 196–210. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  10. 10.
    Meila, M., Shi, J.: A random walks view of spectral segmentation. In: Proc. of 10th International Workshop on Artificial Intelligence and Statistics (AISTATS), pp. 8–11 (2001)Google Scholar
  11. 11.
    Newman, M.E.J.: Detecting community structure in networks. European Physics J. B 38, 321–330 (2004)Google Scholar
  12. 12.
    Ng, A., Jordan, M., Weiss, Y.: On spectral clustering: Analysis and an algorithm. In: Advances in Neural Information Processing Systems 14, pp. 849–856 (2001)Google Scholar
  13. 13.
    Sanchez-Silva, M.: Applicability of Network Clustering Methods for Risk Analysis. In: Topping, B.H.V., Tsompanakis, Y. (eds.) Soft Computing in Civil and Structural Engineering, pp. 283–306. Saxe-Coburg Publications, Stirlingshire (2009)CrossRefGoogle Scholar
  14. 14.
    Shi, T., Belkin, M., Yu, B.: Data spectroscopy: eigenspace of convolution operators and clustering. The Annals of Statistics 37(6B), 3960–3984 (2009)MathSciNetzbMATHCrossRefGoogle Scholar
  15. 15.
    von Luxburg, U.: A tutorial on spectral clustering. J. Statistics and Computing 17(4), 395–416 (2007)CrossRefGoogle Scholar
  16. 16.
    Xia, T., Cao, J., Zhang, Y., Li, J.: On defining affinity graph for spectral clustering through ranking on manifolds. Neurocomputing 72(13-15), 3203–3211 (2008)CrossRefGoogle Scholar
  17. 17.
    Xu, R., Wunsch II, D.: Survey on clustering algorithms. IEEE Trans. on Neural Networks 16(3), 645–678 (2005)CrossRefGoogle Scholar
  18. 18.
    Zelnik-Manor, L., Perona, P.: Self-tuning spectral clustering. In: Proc. of NIPS 2004, pp. 1601–1608 (2004)Google Scholar
  19. 19.
    Zhang, J.: A Clustering Application in Portfolio Management. Lecture Notes in Electrical Engineering, vol. 60, pp. 309–321 (2010)Google Scholar

Copyright information

© IFIP International Federation for Information Processing 2012

Authors and Affiliations

  • Małgorzata Lucińska
    • 1
  • Sławomir T. Wierzchoń
    • 2
    • 3
  1. 1.Kielce University of TechnologyKielcePoland
  2. 2.Institute of Computer Science Polish Academy of SciencesWarsawPoland
  3. 3.University of GdańskGdańskPoland

Personalised recommendations