On the exponential generating function for non-backtracking walks

We derive an explicit formula for the exponential generating function associated with non-backtracking walks around a graph. We study both undirected and directed graphs. Our results allow us to derive computable expressions for non-backtracking versions of network centrality measures based on the matrix exponential. We find that eliminating backtracking walks in this context does not significantly increase the computational expense. We show how the new measures may be interpreted in terms of standard exponential centrality computation on a certain multilayer network. Insights from this block matrix interpretation also allow us to characterize centrality measures arising from general matrix functions. Rigorous analysis on the star graph illustrates the effect of non-backtracking and shows that unwanted localization effects can be eliminated when we restrict to non-backtracking walks. We also investigate the localization issue on synthetic networks.

[1]  Ernesto Estrada,et al.  Communicability in complex networks. , 2007, Physical review. E, Statistical, nonlinear, and soft matter physics.

[2]  A. Terras.  Harmonic Analysis on Symmetric Spaces―Euclidean Space, the Sphere, and the Poincaré Upper Half-Plane , 2013 .

[3]  M. Eiermann,et al.  Implementation of a restarted Krylov subspace method for the evaluation of matrix functions , 2008 .

[4]  Andrei Tarfulea,et al.  An Ihara formula for partially directed graphs , 2009 .

[5]  Albert,et al.  Emergence of scaling in random networks , 1999, Science.

[6]  Peter Grindrod,et al.  The Deformed Graph Laplacian and Its Applications to Network Centrality Analysis , 2018, SIAM J. Matrix Anal. Appl..

[7]  A. Terras,et al.  Zeta Functions of Finite Graphs and Coverings , 1996 .

[8]  N. Alon,et al.  Non-backtracking random walks mix faster , 2006, math/0610550.

[9]  Audry Terras.  What are zeta functions of graphs and what are they good for ? , 2005 .

[10]  P. Bonacich.  Power and Centrality: A Family of Measures , 1987, American Journal of Sociology.

[11]  Yuji Nakatsukasa,et al.  On the stability of computing polynomial roots via confederate linearizations , 2015, Math. Comput..

[12]  Romualdo Pastor-Satorras,et al.  Distinct types of eigenvector localization in networks , 2015, Scientific Reports.

[13]  Alex Townsend,et al.  Vector Spaces of Linearizations for Matrix Polynomials: A Bivariate Polynomial Approach , 2016, SIAM J. Matrix Anal. Appl..

[14]  D. Cvetkovic,et al.  Eigenspaces of graphs: Bibliography , 1997 .

[15]  J. Friedman,et al.  THE NON-BACKTRACKING SPECTRUM OF THE UNIVERSAL COVER OF A GRAPH , 2007, 0712.0192.

[16]  Gábor Szabó,et al.  Structure of complex networks , 2005 .

[17]  Desmond J. Higham,et al.  Network Properties Revealed through Matrix Functions , 2010, SIAM Rev..

[18]  M. Stephanov,et al.  Random Matrices , 2005, hep-ph/0509286.

[19]  Y. Saad.  Analysis of some Krylov subspace approximations to the matrix exponential operator , 1992 .

[20]  Tony White,et al.  The Structure of Complex Networks , 2011 .

[21]  Sergey N. Dorogovtsev,et al.  Localization and Spreading of Diseases in Complex Networks , 2012, Physical review letters.

[22]  Florent Krzakala,et al.  Spectral Clustering of graphs with the Bethe Hessian , 2014, NIPS.

[23]  Tatsuro Kawamoto,et al.  Localized eigenvectors of the non-backtracking matrix , 2015, 1505.07543.

[24]  Desmond J. Higham,et al.  CONTEST: A Controllable Test Matrix Toolbox for MATLAB , 2009, TOMS.

[25]  S. Sodin.  Random matrices, nonbacktracking walks, and orthogonal polynomials , 2007, math-ph/0703043.

[26]  Xiao Zhang,et al.  Localization and centrality in networks , 2014, Physical review. E, Statistical, nonlinear, and soft matter physics.

[27]  Awad H. Al-Mohy,et al.  Computing the Action of the Matrix Exponential, with an Application to Exponential Integrators , 2011, SIAM J. Sci. Comput..

[28]  Kenji Fukumizu,et al.  Graph Zeta Function in the Bethe Free Energy and Loopy Belief Propagation , 2009, NIPS.

[29]  P. Bonacich.  Factoring and weighting approaches to status scores and clique identification , 1972 .

[30]  Hernán A. Makse,et al.  Collective Influence Algorithm to find influencers via optimal percolation in massively large social media , 2016, Scientific Reports.

[31]  Leo Katz,et al.  A new status index derived from sociometric analysis , 1953 .

[32]  Nicholas J. Higham,et al.  Functions of matrices - theory and computation , 2008 .

[33]  Hernán A. Makse,et al.  Influence maximization in complex networks through optimal percolation , 2015, Nature.

[34]  M. Horton.  Ihara zeta functions of digraphs , 2007 .

[35]  Elchanan Mossel,et al.  Spectral redemption in clustering sparse networks , 2013, Proceedings of the National Academy of Sciences.

[36]  Patrick J. Wolfe,et al.  Topology reveals universal features for network comparison , 2017, ArXiv.

[37]  Michele Benzi,et al.  Computation of Generalized Matrix Functions , 2015, SIAM J. Matrix Anal. Appl..

[38]  Peter Grindrod,et al.  Non-backtracking walk centrality for directed networks , 2018, J. Complex Networks.

[39]  Mason A. Porter,et al.  Multilayer networks , 2013, J. Complex Networks.

[40]  Edwin R. Hancock,et al.  Eigenspaces for Graphs , 2002, Int. J. Image Graph..

[41]  A. Terras,et al.  Zeta functions of finite graphs and coverings, III , 1996 .

[42]  Uzy Smilansky,et al.  Quantum chaos on discrete graphs , 2007, 0704.3525.

[43]  Michele Benzi,et al.  The Physics of Communicability in Complex Networks , 2011, ArXiv.

[44]  Federico Poloni,et al.  Duality of matrix pencils, Wong chains and linearizations , 2015 .

[45]  Michele Benzi,et al.  On the Limiting Behavior of Parameter-Dependent Network Centrality Measures , 2013, SIAM J. Matrix Anal. Appl..

[46]  Michele Benzi,et al.  Total communicability as a centrality measure , 2013, J. Complex Networks.