Advertisement

A Jensen-Shannon Divergence Kernel for Directed Graphs

  • Cheng YeEmail author
  • Richard C. Wilson
  • Edwin R. Hancock
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10029)

Abstract

Recently, kernel methods have been widely employed to solve machine learning problems such as classification and clustering. Although there are many existing graph kernel methods for comparing patterns represented by undirected graphs, the corresponding methods for directed structures are less developed. In this paper, to fill this gap in the literature we exploit the graph kernels and graph complexity measures, and present an information theoretic kernel method for assessing the similarity between a pair of directed graphs. In particular, we show how the Jensen-Shannon divergence, which is a mutual information measure that gauges the difference between probability distributions, together with the recently developed directed graph von Neumann entropy, can be used to compute the graph kernel. In the experiments, we show that our kernel method provides an efficient tool for classifying directed graphs with different structures and analyzing real-world complex data.

Keywords

Information theoretic kernel Jensen-Shannon divergence von Neumann entropy 

References

  1. 1.
    Bai, L., Hancock, E.R.: Graph kernels from the Jensen-Shannon divergence. J. Math. Imaging Vis. 47(1–2), 60–69 (2013)CrossRefzbMATHGoogle Scholar
  2. 2.
    Ye, C., Wilson, R.C., Comin, C.H., Costa, L.D.F., Hancock, E.R.: Approximate von Neumann entropy for directed graphs. Phys. Rev. E 89, 052804 (2014)CrossRefGoogle Scholar
  3. 3.
    Cuturi, M., Fukumizu, K., Vert, J.P.: Semigroup kernels on measures. J. Mach. Learn. Res. 6, 1169–1198 (2005)MathSciNetzbMATHGoogle Scholar
  4. 4.
    Martins, A.F.T., Smith, N.A., Xing, E.P., Aguiar, P.M.Q., Figueiredo, M.A.T.: Nonextensive information theoretic kernels on measures. J. Mach. Learn. Res. 10, 935–975 (2009)MathSciNetzbMATHGoogle Scholar
  5. 5.
    Vishwanathan, S.V.N., Schraudolph, N.N., Kondor, R., Borgwardt, K.M.: Graph kernels. J. Mach. Learn. Res. 11, 1201–1242 (2010)MathSciNetzbMATHGoogle Scholar
  6. 6.
    Kashima, H., Tsuda, K., Inokuchi, A.: Marginalized kernels between labeled graphs. In: Proceedings in 20th International Conference on Machine Learning, pp. 321–328(2003)Google Scholar
  7. 7.
    Borgwardt, K.M., Kriegel, H.P.: Shortest-path kernels on graphs. In: Proceedings in the 5th IEEE International Conference on Data Mining, pp. 74–81 (2005)Google Scholar
  8. 8.
    Aziz, F., Wilson, R.C., Hancock, E.R.: Backtrackless walks on a graph. IEEE Trans. Neural Netw. Learn. Syst. 24(6), 977–989 (2013)CrossRefGoogle Scholar
  9. 9.
    Berwanger, D., Gradel, E., Kaiser, L., Rabinovich, R.: Entanglement and the complexity of directed graphs. Theor. Comput. Sci. 463, 2–25 (2012)MathSciNetCrossRefzbMATHGoogle Scholar
  10. 10.
    Escolano, F., Hancock, E.R., Lozano, M.A.: Heat diffusion: thermodynamic depth complexity of networks. Phys. Rev. E 85, 036206 (2012)CrossRefGoogle Scholar
  11. 11.
    Han, L., Escolano, F., Hancock, E.R., Wilson, R.C.: Graph characterizations from von Neumann entropy. Pattern Recogn. Lett. 33, 1958–1967 (2012)CrossRefGoogle Scholar
  12. 12.
    Chung, F.: Laplacians and the Cheeger inequailty for directed graphs. Ann. Comb. 9, 1–19 (2005)MathSciNetCrossRefzbMATHGoogle Scholar
  13. 13.
    Barabási, A.L., Albert, R.: Emergence of scaling in random networks. Science 286, 509–512 (1999)MathSciNetCrossRefzbMATHGoogle Scholar
  14. 14.
    Battiston, S., Caldarelli, G.: Systemic risk in financial networks. J. Financ. Manag. Mark. Inst. 1, 129–154 (2013)Google Scholar
  15. 15.
    Schölkopf, B., Smola, A., Müller, K.B.: Nonlinear component analysis as a kernel eigenvalue problem. Neural Comput. 10, 1299–1319 (1998)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2016

Authors and Affiliations

  • Cheng Ye
    • 1
    Email author
  • Richard C. Wilson
    • 1
  • Edwin R. Hancock
    • 1
  1. 1.Department of Computer ScienceUniversity of YorkYorkUK

Personalised recommendations