Sharp Bounds on Random Walk Eigenvalues via Spectral Embedding

Spectral embedding of graphs uses the top k non-trivial eigenvectors of the random walk matrix to embed the graph into R^k. The primary use of this embedding has been for practical spectral clustering algorithms [SM00,NJW02]. Recently, spectral embedding was studied from a theoretical perspective to prove higher order variants of Cheeger's inequality [LOT12,LRTV12]. We use spectral embedding to provide a unifying framework for bounding all the eigenvalues of graphs. For example, we show that for any finite graph with n vertices and all k >= 2, the k-th largest eigenvalue is at most 1-Omega(k^3/n^3), which extends the only other such result known, which is for k=2 only and is due to [LO81]. This upper bound improves to 1-Omega(k^2/n^2) if the graph is regular. We generalize these results, and we provide sharp bounds on the spectral measure of various classes of graphs, including vertex-transitive graphs and infinite graphs, in terms of specific graph parameters like the volume growth. As a consequence, using the entire spectrum, we provide (improved) upper bounds on the return probabilities and mixing time of random walks with considerably shorter and more direct proofs. Our work introduces spectral embedding as a new tool in analyzing reversible Markov chains. Furthermore, building on [Lyo05], we design a local algorithm to approximate the number of spanning trees of massive graphs.

[1]  V. Climenhaga.  Markov chains and mixing times , 2013 .

[2]  Ramin Javadi,et al.  On nodal domains and higher-order Cheeger inequalities of finite reversible Markov processes , 2012 .

[3]  James R. Lee,et al.  Multi-way spectral partitioning and higher-order cheeger inequalities , 2011, STOC '12.

[4]  Prasad Raghavendra,et al.  Many sparse cuts via higher eigenvalues , 2011, STOC '12.

[5]  C. Pittet,et al.  Spectral properties of a class of random walks on locally finite groups , 2011, 1102.1952.

[6]  James R. Lee,et al.  Metric Uniformization and Spectral Bounds for Graphs , 2010, ArXiv.

[7]  Gábor Elek,et al.  Parameter testing in bounded degree graphs of subexponential growth , 2007, Random Struct. Algorithms.

[8]  Hua Bai.  The Grone-Merris Conjecture , 2009, 0911.2172.

[9]  Roman Sauer,et al.  Spectral distribution and L2-isoperimetric profile of Laplace operators on groups , 2009, 0901.0271.

[10]  A. Cayley.  A theorem on trees , 2009 .

[11]  Ulrike von Luxburg,et al.  A tutorial on spectral clustering , 2007, Stat. Comput..

[12]  Avi Wigderson,et al.  Pairwise Independence and Derandomization , 2006, Found. Trends Theor. Comput. Sci..

[13]  Ravi Montenegro,et al.  Mathematical Aspects of Mixing Times in Markov Chains , 2006, Found. Trends Theor. Comput. Sci..

[14]  M. Barlow,et al.  Characterization of sub‐Gaussian heat kernel estimates on strongly recurrent graphs , 2005 .

[15]  P. Tetali,et al.  Mixing Time Bounds via the Spectral Profile , 2005, math/0505690.

[16]  R. Lyons.  Asymptotic Enumeration of Spanning Trees , 2002, Combinatorics, Probability and Computing.

[17]  Y. Peres,et al.  Evolving sets, mixing and heat kernel bounds , 2003, math/0305349.

[18]  Victor Reiner,et al.  Shifted simplicial complexes are Laplacian integral , 2002 .

[19]  Thierry Coulhon,et al.  Manifolds and graphs with slow heat kernel decay , 2001 .

[20]  Michael I. Jordan,et al.  On Spectral Clustering: Analysis and an algorithm , 2001, NIPS.

[21]  Thierry Coulhon,et al.  A geometric approach to on-diagonal heat kernel lower bounds on groups , 2001 .

[22]  Jitendra Malik,et al.  Normalized cuts and image segmentation , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[23]  Charles Delorme,et al.  Eigenvalues, eigenspaces and distances to subsets , 1997, Discret. Math..

[24]  Thierry Coulhon,et al.  Ultracontractivity and Nash Type Inequalities , 1996 .

[25]  Uriel Feige,et al.  Random Walks on Regular and Irregular Graphs , 1996, SIAM J. Discret. Math..

[26]  Joel Friedman,et al.  Minimum higher eigenvalues of Laplacians on graphs , 1996 .

[27]  Laurent Saloff-Coste,et al.  Isoperimetric Inequalities and Decay of Iterated Kernels for Almost-transitive Markov Chains , 1995, Combinatorics, Probability and Computing.

[28]  Shing-Tung Yau,et al.  Eigenvalues of Graphs and Sobolev Inequalities , 1995, Combinatorics, Probability and Computing.

[29]  Russell Merris,et al.  The Laplacian Spectrum of a Graph II , 1994, SIAM J. Discret. Math..

[30]  M THEORE,et al.  Moderate Growth and Random Walk on Finite Groups , 1994 .

[31]  P. Diaconis,et al.  Comparison Techniques for Random Walk on Finite Groups , 1993 .

[32]  M. Gromov,et al.  Von Neumann spectra near zero , 1991 .

[33]  László Babai,et al.  Local expansion of vertex-transitive graphs and random generation in finite groups , 1991, STOC '91.

[34]  V. Sunder,et al.  The Laplacian spectrum of a graph , 1990 .

[35]  M. Shubin,et al.  Spectrum distribution function and variational principle for automorphic operators on hyperbolic space , 1989 .

[36]  V. Trofimov,et al.  GRAPHS WITH POLYNOMIAL GROWTH , 1985 .

[37]  A. Odlyzko,et al.  Bounds for eigenvalues of certain stochastic matrices , 1981 .

[38]  Peter Li.  Eigenvalue estimates on homogeneous manifolds , 1980 .

[39]  J. Moon,et al.  On the diameter of a graph. , 1965 .

[40]  M. Bálek,et al.  Large Networks and Graph Limits , 2022 .