Changepoint Detection over Graphs with the Spectral Scan Statistic

We consider the change-point detection problem of deciding, based on noisy measurements, whether an unknown signal over a given graph is constant or is instead piecewise constant over two connected induced subgraphs of relatively low cut size. We analyze the corresponding generalized likelihood ratio (GLR) statistics and relate it to the problem of finding a sparsest cut in a graph. We develop a tractable relaxation of the GLR statistic based on the combinatorial Laplacian of the graph, which we call the spectral scan statistic, and analyze its properties. We show how its performance as a testing procedure depends directly on the spectrum of the graph, and use this result to explicitly derive its asymptotic properties on few significant graph topologies. Finally, we demonstrate both theoretically and by simulations that the spectral scan statistic can outperform naive testing procedures based on edge thresholding and $\chi^2$ testing.

[1]  Louis L. Scharf,et al.  Matched subspace detectors , 1994, IEEE Trans. Signal Process..

[2]  Oscar Rojo,et al.  The spectra of the adjacency matrix and Laplacian matrix for some balanced trees , 2005 .

[3]  R. Merris.  Laplacian graph eigenvectors , 1998 .

[4]  Aarti Singh,et al.  Identifying graph-structured activation patterns in networks , 2010, NIPS.

[5]  Gordon F. Royle,et al.  Algebraic Graph Theory , 2001, Graduate texts in mathematics.

[6]  Alessandro Rinaldo,et al.  Sparsistency of the Edge Lasso over Graphs , 2012, AISTATS.

[7]  Bryan L. Shader,et al.  Tight bounds on the algebraic connectivity of a balanced binary tree , 1999 .

[8]  Christos Faloutsos,et al.  Kronecker Graphs: An Approach to Modeling Networks , 2008, J. Mach. Learn. Res..

[9]  Frank Thomson Leighton,et al.  An approximate max-flow min-cut theorem for uniform multicommodity flow problems with applications to approximation algorithms , 1988, [Proceedings 1988] 29th Annual Symposium on Foundations of Computer Science.

[10]  Igor V. Nikiforov,et al.  Asymptotically Uniformly Minimax Detection and Isolation in Network Monitoring , 2012, IEEE Transactions on Signal Processing.

[11]  Farhad Shahrokhi,et al.  Sparsest cuts and bottlenecks in graphs , 1990, Discret. Appl. Math..

[12]  Michalis Faloutsos,et al.  On power-law relationships of the Internet topology , 1999, SIGCOMM '99.

[13]  A. Wald.  Tests of statistical hypotheses concerning several parameters when the number of observations is large , 1943 .

[14]  David B. Shmoys,et al.  Cut problems and their application to divide-and-conquer , 1996 .

[15]  M. Talagrand.  The Generic Chaining , 2005 .

[16]  E. Candès,et al.  Detection of an anomalous cluster in a network , 2010, 1001.3209.

[17]  Alfred O. Hero,et al.  Optimal simultaneous detection and estimation under a false alarm constraint , 1995, IEEE Trans. Inf. Theory.

[18]  Yu. I. Ingster,et al.  Nonparametric Goodness-of-Fit Testing Under Gaussian Models , 2002 .

[19]  S. Dudoit,et al.  Gains in Power from Structured Two-Sample Tests of Means on Graphs , 2010, 1009.5173.

[20]  M. Ledoux.  The concentration of measure phenomenon , 2001 .

[21]  Aarti Singh,et al.  Subspace detection of high-dimensional vectors using compressive sampling , 2012, 2012 IEEE Statistical Signal Processing Workshop (SSP).

[22]  Sharon L. Milgram,et al.  The Small World Problem , 1967 .

[23]  Mitra Fouladirad,et al.  Optimal statistical fault detection with nuisance parameters , 2005, Autom..

[24]  C. Borell.  The Brunn-Minkowski inequality in Gauss space , 1975 .

[25]  E. Candès,et al.  Searching for a trail of evidence in a maze , 2007, math/0701668.

[26]  Peter Kulchyski.  and , 2015 .

[27]  Richard Peng,et al.  Faster Spectral Sparsification and Numerical Algorithms for SDD Matrices , 2012, ACM Trans. Algorithms.

[28]  Stephen E. Fienberg,et al.  Testing Statistical Hypotheses , 2005 .

[29]  Avi Ostfeld,et al.  The Battle of the Water Sensor Networks (BWSN): A Design Challenge for Engineers and Algorithms , 2008 .

[30]  Luc Devroye,et al.  Combinatorial Testing Problems , 2009, 0908.3437.

[31]  Andrew B. Kahng,et al.  New spectral methods for ratio cut partitioning and clustering , 1991, IEEE Trans. Comput. Aided Des. Integr. Circuits Syst..

[32]  Mitra Fouladirad,et al.  Optimal fault detection with nuisance parameters and a general covariance matrix , 2008 .

[33]  W. N. Burnette,et al.  "Western blotting": electrophoretic transfer of proteins from sodium dodecyl sulfate--polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. , 1981, Analytical biochemistry.

[34]  Igor V. Nikiforov,et al.  Non-Bayesian Detection and Detectability of Anomalies From a Few Noisy Tomographic Projections , 2007, IEEE Transactions on Signal Processing.

[35]  M. Fiedler.  Eigenvectors of acyclic matrices , 1975 .

[36]  P. Cochat,et al.  Et al , 2008, Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.

[37]  Christos Faloutsos,et al.  Scalable modeling of real graphs using Kronecker multiplication , 2007, ICML '07.

[38]  Xiaoming Huo,et al.  Near-optimal detection of geometric objects by fast multiscale methods , 2005, IEEE Transactions on Information Theory.

[39]  Oscar Rojo,et al.  The spectrum of the Laplacian matrix of a balanced binary tree , 2002 .

[40]  I. Ibragimov,et al.  Norms of Gaussian sample functions , 1976 .

[41]  F. Chung.  Discrete isoperimetric inequalities , 1996 .

[42]  Satish Rao,et al.  Expander flows, geometric embeddings and graph partitioning , 2004, STOC '04.