Fast Eigenspace Approximation using Random Signals

We focus in this work on the estimation of the first $k$ eigenvectors of any graph Laplacian using filtering of Gaussian random signals. We prove that we only need $k$ such signals to be able to exactly recover as many of the smallest eigenvectors, regardless of the number of nodes in the graph. In addition, we address key issues in implementing the theoretical concepts in practice using accurate approximated methods. We also propose fast algorithms both for eigenspace approximation and for the determination of the $k$th smallest eigenvalue $\lambda_k$. The latter proves to be extremely efficient under the assumption of locally uniform distribution of the eigenvalue over the spectrum. Finally, we present experiments which show the validity of our method in practice and compare it to state-of-the-art methods for clustering and visualization both on synthetic small-scale datasets and larger real-world problems of millions of nodes. We show that our method allows a better scaling with the number of nodes than all previous methods while achieving an almost perfect reconstruction of the eigenspace formed by the first $k$ eigenvectors.

[1]  P. Vandergheynst,et al.  Accelerated filtering on graphs using Lanczos method , 2015, 1509.04537.

[2]  Ping Tak Peter Tang,et al.  FEAST As A Subspace Iteration Eigensolver Accelerated By Approximate Spectral Projection , 2013, SIAM J. Matrix Anal. Appl..

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

[4]  Michael W. Mahoney,et al.  Implementing regularization implicitly via approximate eigenvector computation , 2010, ICML.

[5]  R. C. Ward,et al.  High performance parallel approximate eigensolver for real symmetric matrices , 2005 .

[6]  Christos Boutsidis,et al.  Near Optimal Column-Based Matrix Reconstruction , 2011, 2011 IEEE 52nd Annual Symposium on Foundations of Computer Science.

[7]  Heng Tao Shen,et al.  Principal Component Analysis , 2009, Encyclopedia of Biometrics.

[8]  Edoardo Di Napoli,et al.  Efficient estimation of eigenvalue counts in an interval , 2013, Numer. Linear Algebra Appl..

[9]  Roman Vershynin,et al.  Introduction to the non-asymptotic analysis of random matrices , 2010, Compressed Sensing.

[10]  Jure Leskovec,et al.  Defining and evaluating network communities based on ground-truth , 2012, Knowledge and Information Systems.

[11]  Jack Dongarra,et al.  Templates for the Solution of Algebraic Eigenvalue Problems , 2000, Software, environments, tools.

[12]  Geoffrey E. Hinton,et al.  Stochastic Neighbor Embedding , 2002, NIPS.

[13]  Pierre Vandergheynst,et al.  GSPBOX: A toolbox for signal processing on graphs , 2014, ArXiv.

[14]  Geoffrey E. Hinton,et al.  Visualizing Data using t-SNE , 2008 .

[15]  V. Marčenko,et al.  DISTRIBUTION OF EIGENVALUES FOR SOME SETS OF RANDOM MATRICES , 1967 .

[16]  C. Lanczos An iteration method for the solution of the eigenvalue problem of linear differential and integral operators , 1950 .

[17]  Upamanyu Madhow,et al.  Compressive spectral embedding: sidestepping the SVD , 2015, NIPS.

[18]  Jingzhou Liu,et al.  Visualizing Large-scale and High-dimensional Data , 2016, WWW.

[19]  Mikhail Belkin,et al.  Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering , 2001, NIPS.

[20]  D. Calvetti,et al.  AN IMPLICITLY RESTARTED LANCZOS METHOD FOR LARGE SYMMETRIC EIGENVALUE PROBLEMS , 1994 .

[21]  U. Feige,et al.  Spectral Graph Theory , 2015 .

[22]  M. Cugmas,et al.  On comparing partitions , 2015 .

[23]  M E J Newman,et al.  Finding and evaluating community structure in networks. , 2003, Physical review. E, Statistical, nonlinear, and soft matter physics.

[24]  Christos Boutsidis,et al.  Spectral Clustering via the Power Method - Provably , 2013, ICML.

[25]  Danny C. Sorensen,et al.  Implicit Application of Polynomial Filters in a k-Step Arnoldi Method , 1992, SIAM J. Matrix Anal. Appl..

[26]  C. Chui,et al.  Article in Press Applied and Computational Harmonic Analysis a Randomized Algorithm for the Decomposition of Matrices , 2022 .

[27]  Pierre Vandergheynst,et al.  Compressive Spectral Clustering , 2016, ICML.

[28]  W. Arnoldi The principle of minimized iterations in the solution of the matrix eigenvalue problem , 1951 .

[29]  Pierre Vandergheynst,et al.  Vertex-Frequency Analysis on Graphs , 2013, ArXiv.

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

[31]  Pierre Vandergheynst,et al.  Wavelets on Graphs via Spectral Graph Theory , 2009, ArXiv.