Universality of covariance matrices

DOI10.1214/13-AAP939MaRDI QIDQ2454401zbMATH OpenFDO

Authors Natesh S. Pillai, Jun Yin

Publication date 13 June 2014

Published in The Annals of Applied Probability (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1110.2501, https://projecteuclid.org/euclid.aoap/1398258093

zbMATH Keywords

Stieltjes transformcovariance matrixuniversalityWishart matrixDyson Brownian motionTracy-Widom lawMarcenko-Pastur lawasymptotic distribution of the eigenvaluesextremal singular valueslocal statistics of eigenvalues

Mathematics Subject Classification ID

Eigenvalues, singular values, and eigenvectors (15A18) Random matrices (algebraic aspects) (15B52) Random matrices (probabilistic aspects) (60B20) Brownian motion (60J65) Disordered systems (random Ising models, random Schrödinger operators, etc.) in equilibrium statistical mechanics (82B44)

Abstract: In this paper we prove the universality of covariance matrices of the form $H_{NimesN}=X^{dagger}X$ where $X$ is an $MimesN$ rectangular matrix with independent real valued entries $x_{ij}$ satisfying $mathbbE{x}_{ij}=0$ and $mathbbE{x}_{ij}^2=frac1M$, $N$, $Moinfty$. Furthermore it is assumed that these entries have sub-exponential tails or sufficiently high number of moments. We will study the asymptotics in the regime $N/M=d_{N}in(0,infty),li{m}_{Noinfty}{d}_{N}eq0,infty$. Our main result is the edge universality of the sample covariance matrix at both edges of the spectrum. In the case $li{m}_{Noinfty}{d}_N=1$, we only focus on the largest eigenvalue. Our proof is based on a novel version of the Green function comparison theorem for data matrices with dependent entries. En route to proving edge universality, we establish that the Stieltjes transform of the empirical eigenvalue distribution of $H$ is given by the Marcenko-Pastur law uniformly up to the edges of the spectrum with an error of order $(Neta{)}^{-1}$ where $eta$ is the imaginary part of the spectral parameter in the Stieltjes transform. Combining these results with existing techniques we also show bulk universality of covariance matrices. All our results hold for both real and complex valued entries.

Recommendations

• Random covariance matrices: universality of local statistics of eigenvalues
• Random covariance matrices: universality of local statistics of eigenvalues up to the edge
• On universality of bulk local regime of the Hermitian sample covariance matrices
• Universality in the bulk of the spectrum for complex sample covariance matrices
• Universality for the largest eigenvalue of sample covariance matrices with general population

Cites work

• scientific article; zbMATH DE number 5204610 (Why is no real title available?)
• A Brownian-Motion Model for the Eigenvalues of a Random Matrix
• A comment on the Wigner-Dyson-Mehta bulk universality conjecture for Wigner matrices
• A necessary and sufficient condition for edge universality of Wigner matrices
• A note on the largest eigenvalue of a large dimensional sample covariance matrix
• A note on universality of the distribution of the largest eigenvalues in certain sample covariance matrices
• A universality result for the smallest eigenvalues of certain sample covariance matrices
• Application of Random Matrix Theory to Multivariate Statistics
• Bulk universality for Wigner matrices
• Bulk universality for generalized Wigner matrices
• DISTRIBUTION OF EIGENVALUES FOR SOME SETS OF RANDOM MATRICES
• Edge universality of correlation matrices
• Eigenvector distribution of Wigner matrices
• Explaining the single factor bias of arbitrage pricing models in finite samples
• Limit of the smallest eigenvalue of a large dimensional sample covariance matrix
• Local Marchenko-Pastur law at the hard edge of sample covariance matrices
• Local circular law for random matrices
• Multivariate analysis and Jacobi ensembles: largest eigenvalue, Tracy-Widom limits and rates of convergence
• On the distribution of the largest eigenvalue in principal components analysis
• On the limit of the largest eigenvalue of the large dimensional sample covariance matrix
• Random covariance matrices: universality of local statistics of eigenvalues
• Random covariance matrices: universality of local statistics of eigenvalues up to the edge
• Random matrices: The distribution of the smallest singular values
• Random matrices: Universality of local eigenvalue statistics up to the edge
• Random matrices: universality of local eigenvalue statistics
• Rate of convergence in probability to the Marchenko-Pastur law
• Rigidity of eigenvalues of generalized Wigner matrices
• Spectral statistics of Erdős-Rényi graphs II: eigenvalue spacing and the extreme eigenvalues
• Spectral statistics of Erdős-Rényi graphs. I: Local semicircle law
• Testing hypotheses about the number of factors in large factor models
• The distribution and moments of the smallest eigenvalue of a random matrix of Wishart type
• The isotropic semicircle law and deformation of Wigner matrices
• The local relaxation flow approach to universality of the local statistics for random matrices
• Tracy-Widom law for the extreme eigenvalues of sample correlation matrices
• Universality for generalized Wigner matrices with Bernoulli distribution
• Universality in the bulk of the spectrum for complex sample covariance matrices
• Universality of local eigenvalue statistics for some sample covariance matrices
• Universality of random matrices and local relaxation flow
• Universality results for the largest eigenvalues of some sample covariance matrix ensembles

Cited in

(75)

• Order Determination for Spiked Type Models
• Anisotropic local laws for random matrices
• Large complex correlated Wishart matrices: fluctuations and asymptotic independence at the edges
• Goodness-of-fit test for latent block models
• scientific article; zbMATH DE number 3927698 (Why is no real title available?)
• Quantitative Tracy-Widom laws for the largest eigenvalue of generalized Wigner matrices
• Contiguity under high-dimensional Gaussianity with applications to covariance testing
• Limit theorems for the counting function of eigenvalues up to edge in covariance matrices
• Free energy of bipartite spherical Sherrington-Kirkpatrick model
• Local Marchenko–Pastur law at the hard edge of the sample covariance ensemble
• A spectral approach to Hebbian-like neural networks
• On universality of bulk local regime of the Hermitian sample covariance matrices
• A necessary and sufficient condition for edge universality at the largest singular values of covariance matrices
• A necessary and sufficient condition for edge universality of Wigner matrices
• Spiked sample covariance matrices with possibly multiple bulk components
• Universal halting times in optimization and machine learning
• High-dimensional dynamics of generalization error in neural networks
• Wegner estimate and upper bound on the eigenvalue condition number of non-Hermitian random matrices
• Eigenvalue variance bounds for covariance matrices
• Convergence rate to the Tracy-Widom laws for the largest eigenvalue of Wigner matrices
• Quantitative universality for the largest eigenvalue of sample covariance matrices
• Averaging fluctuations in resolvents of random band matrices
• Spectral asymptotic expansion of Wishart matrices with exploding moments
• Local circular law for random matrices
• The local circular law. II: The edge case
• Tracy-Widom limit for Kendall's tau
• The Tracy-Widom law for the largest eigenvalue of F type matrices
• The spectral norm of random inner-product kernel matrices
• Bounds for the Stieltjes transform and the density of states of Wigner matrices
• Random band matrices in the delocalized phase. III: Averaging fluctuations
• On the principal components of sample covariance matrices
• Some strong convergence theorems for eigenvalues of general sample covariance matrices
• Sparse recovery from extreme eigenvalues deviation inequalities
• Small deviation estimates for the largest eigenvalue of Wigner matrices
• On delocalization of eigenvectors of random non-Hermitian matrices
• Eigenvectors of random matrices: A survey
• A goodness-of-fit test on the number of biclusters in a relational data matrix
• Isotropic self-consistent equations for mean-field random matrices
• Covariance matrices of self-affine measures
• Convergence rate to the Tracy-Widom laws for the largest eigenvalue of sample covariance matrices
• Asymptotic freeness for rectangular random matrices and large deviations for sample covariance matrices with sub-Gaussian tails
• Spiked separable covariance matrices and principal components
• Convergence of eigenvector empirical spectral distribution of sample covariance matrices
• Sampling without replacement from a high-dimensional finite population
• Edge universality of separable covariance matrices
• High dimensional deformed rectangular matrices with applications in matrix denoising
• The conjugate gradient algorithm on well-conditioned Wishart matrices is almost deterministic
• Norm convergence rate for multivariate quadratic polynomials of Wigner matrices
• Sample canonical correlation coefficients of high-dimensional random vectors: Local law and Tracy–Widom limit
• Extremal eigenvalues of sample covariance matrices with general population
• Delocalization and diffusion profile for random band matrices
• Universality of the least singular value for sparse random matrices
• Tracy-Widom limit for the largest eigenvalue of high-dimensional covariance matrices in elliptical distributions
• On the convergence of the extremal eigenvalues of empirical covariance matrices with dependence
• Edge universality of correlation matrices
• Spectrum of high-dimensional sample covariance and related matrices: a selective review
• On estimation in the reduced-rank regression with a large number of responses and predictors
• Universality for Eigenvalue Algorithms on Sample Covariance Matrices
• Singular vector distribution of sample covariance matrices
• Data-driven optimal shrinkage of singular values under high-dimensional noise with separable covariance structure with application
• Fluctuations of the free energy of the spherical Sherrington-Kirkpatrick model
• Linear spectral statistics of eigenvectors of anisotropic sample covariance matrices
• Universality for the largest eigenvalue of sample covariance matrices with general population
• Tail bounds for gaps between eigenvalues of sparse random matrices
• Edge statistics of large dimensional deformed rectangular matrices
• Large sample correlation matrices: a comparison theorem and its applications
• Tracy-Widom at each edge of real covariance and MANOVA estimators
• Limiting distributions for eigenvalues of sample correlation matrices from heavy-tailed populations
• Gaussian fluctuations in complex sample covariance matrices
• Local eigenvalue density for general MANOVA matrices
• Local law and Tracy-Widom limit for sparse sample covariance matrices
• Convergence rates of eigenvector empirical spectral distribution of large dimensional sample covariance matrix
• No outliers in the spectrum of the product of independent non-Hermitian random matrices with independent entries
• Singular vector and singular subspace distribution for the matrix denoising model
• Random covariance matrices: universality of local statistics of eigenvalues up to the edge

This page was built for publication: Universality of covariance matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2454401)