Abstract
Kernels and, broadly speaking, similarity measures on graphs are extensively used in graph-based unsupervised and semi-supervised learning algorithms as well as in the link prediction problem. We analytically study proximity and distance properties of various kernels and similarity measures on graphs. This can potentially be useful for recommending the adoption of one or another similarity measure in a machine learning method. Also, we numerically compare various similarity measures in the context of spectral clustering and observe that normalized heat-type similarity measures with log modification generally perform the best.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
M. Saerens [36] has remarked that a more suitable name could be Neumann diffusion kernel, referring to the Neumann series \(\sum _{k=0}^\infty T^k\) (where T is an operator) named after Carl Gottfried Neumann, while a connection of that to John von Neumann is not obvious (the concept of von Neumann kernel in group theory is essentially different).
- 2.
In fact, L. Katz considered \(\sum _{k=1}^\infty (\alpha W)^k.\)
- 3.
For the properties of M-matrices, we refer to [29].
- 4.
- 5.
On various alternative versions of the triangle inequality, we refer to [17].
References
Avrachenkov, K., Mishenin, A., Gonçalves, P., Sokol, M.: Generalized optimization framework for graph-based semi-supervised learning. In: Proceedings of the 2012 SIAM International Conference on Data Mining, pp. 966–974 (2012)
Avrachenkov, K., Gonçalves, P., Sokol, M.: On the choice of kernel and labelled data in semi-supervised learning methods. In: Bonato, A., Mitzenmacher, M., Prałat, P. (eds.) WAW 2013. LNCS, vol. 8305, pp. 56–67. Springer, Cham (2013). doi:10.1007/978-3-319-03536-9_5
Avrachenkov, K., Chebotarev, P., Mishenin, A.: Semi-supervised learning with regularized Laplacian. Optim. Methods Softw. 32(2), 222–236 (2017)
Avrachenkov, K., van der Hofstad, R., Sokol, M.: Personalized PageRank with node-dependent restart. In: Proceedings of International Workshop on Algorithms and Models for the Web-Graph, pp. 23–33 (2014)
Backstrom, L., Leskovec, J.: Supervised random walks: predicting and recommending links in social networks. Proc. ACM WSDM 2011, 635–644 (2011)
Boley, D., Ranjan, G., Zhang, Z.L.: Commute times for a directed graph using an asymmetric Laplacian. Linear Algebra Appl. 435(2), 224–242 (2011)
Chapelle, O., Schölkopf, B., Zien, A.: Semi-Supervised Learning. MIT Press, Cambridge (2006)
Chebotarev, P.: The graph bottleneck identity. Adv. Appl. Math. 47(3), 403–413 (2011)
Chebotarev, P.: A class of graph-geodetic distances generalizing the shortest-path and the resistance distances. Discrete Appl. Math. 47(3), 403–413 (2011)
Chebotarev, P.: The walk distances in graphs. Discrete Appl. Math. 160(10–11), 1484–1500 (2012)
Chebotarev, P. Yu., Shamis, E.V.: On the proximity measure for graph vertices provided by the inverse Laplacian characteristic matrix. In: Abstracts of the conference “Linear Algebra and its Application”, 10–12 June 1995, The Institute of Mathematics and its Applications, in conjunction with the Manchester Center for Computational Mathematics, Manchester, UK (pp. 6–7), URL http://www.ma.man.ac.uk/higham/laa95/abstracts.ps (1995)
Chebotarev, P.Y., Shamis, E.V.: The matrix-forest theorem and measuring relations in small social groups. Autom. Remote Control 58(9), 1505–1514 (1997)
Chebotarev, P.Y., Shamis, E.V.: On a duality between metrics and \(\varSigma \)-proximities. Autom. Remote Control 59(4), 608–612 (1998)
Chebotarev, P.Y., Shamis, E.V.: On proximity measures for graph vertices. Autom. Remote Control 59(10), 1443–1459 (1998)
Chung, F.: Spectral graph theory, vol. 92. American Math. Soc. (1997)
Chung, F.: The heat kernel as the pagerank of a graph. Proc. Natl. Acad. Sci. 104(50), 19735–19740 (2007)
Deza, M., Chebotarev, P.: Protometrics. arXiv preprint arXiv:1112.4829 (2011)
Dhillon, I.S., Fan, J., Guan, Y.: Efficient clustering of very large document collections. Data Min. sci. Eng. Appl. 2, 357–381 (2001)
Dhillon, I.S., Guan, Y., Kulis, B.: Kernel k-means: spectral clustering and normalized cuts. Proc. ACM KDD 2004, 551–556 (2004)
Estrada, E., Hatano, N.: Statistical-mechanical approach to subgraph centrality in complex networks. Chem. Phys. Lett. 439, 247–251 (2007)
Estrada, E., Hatano, N.: Communicability in complex networks. Phys. Rev. E 77(3), 036111 (2008)
Estrada, E., Silver, G.: Accounting for the role of long walks on networks via a new matrix function. J. Math. Anal. Appl. 449, 1581–1600 (2017)
Fouss, F., Yen L., Pirotte, A., Saerens, M.: An experimental investigation of graph kernels on a collaborative recommendation task. In: Proceedings of the Sixth International Conference on Data Mining (ICDM 2006), pp. 863–868, IEEE (2006)
Fouss, F., Saerens, M., Shimbo, M.: Algorithms and Models for Network Data and Link Analysis. Cambridge University Press, Cambridge (2016)
Horn, R.A., Johnson, C.R.: Matrix Analysis, 2nd edn. Cambridge University Press, Cambridge (2013)
Ivashkin, V., Chebotarev, P.: Do logarithmic proximity measures outperform plain ones in graph clustering? In: Kalyagin, V.A., et al. (eds.) Models, Algorithms and Technologies for Network Analysis. Springer Proceedings in Mathematics & Statistics, vol. 197, pp. 87–105. Springer, Cham (2017)
Jacobsen, K., Tien, J.: A generalized inverse for graphs with absorption. arXiv preprint arXiv:1611.02233 (2016)
Katz, L.: A new status index derived from sociometric analysis. Psychometrika 18(1), 39–43 (1953)
Kirkland, S.J., Neumann, M.: Group Inverses of M-matrices and Their Applications. CRC Press, Boca Raton (2012)
Kivimäki, I., Shimbo, M., Saerens, M.: Developments in the theory of randomized shortest paths with a comparison of graph node distances. Phys. A 393, 600616 (2014)
Kondor, R.I., Lafferty, J.: Diffusion kernels on graphs and other discrete input spaces. In: Proceedings of ICML, pp. 315–322 (2002)
Lenart, C.: A generalized distance in graphs and centered partitions. SIAM J. Discrete Math. 11(2), 293–304 (1998)
Liben-Nowell, D., Kleinberg, J.: The link-prediction problem for social networks. J. Assoc. Inform. Sci. Technol. 58(7), 1019–1031 (2007)
Schoenberg, I.J.: Remarks to Maurice Fréchet’s article “Sur la définition axiomatique d’une classe d’espace distanciés vectoriellement applicable sur l’espace de Hilbert”. Ann. Math. 36(3), 724–732 (1935)
Schoenberg, I.J.: Metric spaces and positive definite functions. Trans. Am. Math. Soc. 44(3), 522–536 (1938)
Saerens, M.: Personal communication
Kandola, J., Shawe-Taylor, J., Cristianini, N.: Learning semantic similarity. In: Neural Information Processing Systems 15 (NIPS 2015). MIT Press (2002)
Shawe-Taylor, J., Cristianini, N.: Kernel Methods for Pattern Analysis. Cambridge University Press, Cambridge (2004)
Smola, A.J., Kondor, R.: Kernels and regularization on graphs. In: Learning Theory and Kernel Machines, pp. 144–158 (2003)
Sommer, F., Fouss, F., Saerens, M.: Comparison of graph node distances on clustering tasks. In: Villa, A.E.P., Masulli, P., Pons Rivero, A.J. (eds.) ICANN 2016. LNCS, vol. 9886, pp. 192–201. Springer, Cham (2016). doi:10.1007/978-3-319-44778-0_23
Vishwanathan, S.V.N., Schraudolph, N.N., Kondor, R., Borgwardt, K.M.: Graph kernels. J. Mach. Learn. Res. 11, 1201–1242 (2010)
von Luxburg, U.: A tutorial on spectral clustering. Stat. Comput. 17(4), 395–416 (2007)
Zhou, D., Schölkopf, B., Hofmann, T.: Semi-supervised learning on directed graphs. In: Proceeedings of NIPS, pp. 1633–1640 (2004)
Müller, K.-R., Mika, S., Rätsch, G., Tsuda, K., Schölkopf, B.: An Introduction to kernel-based learning algorithms. IEEE Trans. Neural Networks 12(2), 181–202 (2001)
Acknowledgements
The work of KA and DR was supported by the joint Bell Labs Inria ADR “Network Science” and by UCA-JEDI Idex Grant “HGRAPHS”, and the work of PC was supported by the Russian Science Foundation (project no.16-11-00063 granted to IRE RAS).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Avrachenkov, K., Chebotarev, P., Rubanov, D. (2017). Kernels on Graphs as Proximity Measures. In: Bonato, A., Chung Graham, F., Prałat, P. (eds) Algorithms and Models for the Web Graph. WAW 2017. Lecture Notes in Computer Science(), vol 10519. Springer, Cham. https://doi.org/10.1007/978-3-319-67810-8_3
Download citation
DOI: https://doi.org/10.1007/978-3-319-67810-8_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-67809-2
Online ISBN: 978-3-319-67810-8
eBook Packages: Computer ScienceComputer Science (R0)