Entrywise limit theorems for eigenvectors of signal-plus-noise matrix models with weak signals

DOI10.3150/23-BEJ1602arXiv2106.09840OpenAlexW4388513805MaRDI QIDQ6178566FDOQ6178566

Authors: Fangzheng Xie 

Publication date: 16 January 2024

Published in: Bernoulli (Search for Journal in Brave)

Abstract: We establish a finite-sample Berry-Esseen theorem for the entrywise limits of the eigenvectors for a broad collection of signal-plus-noise random matrix models under challenging weak signal regimes. The signal strength is characterized by a scaling factor $h{o}_n$ through $nh{o}_n$, where $n$ is the dimension of the random matrix, and we allow $nh{o}_n$ to grow at the rate of $logn$. The key technical contribution is a sharp finite-sample entrywise eigenvector perturbation bound. The existing error bounds on the two-to-infinity norms of the higher-order remainders are not sufficient when $nh{o}_n$ is proportional to $logn$. We apply the general entrywise eigenvector analysis results to the symmetric noisy matrix completion problem, random dot product graphs, and two subsequent inference tasks for random graphs: the estimation of pure nodes in mixed membership stochastic block models and the hypothesis testing of the equality of latent positions in random graphs.

Full work available at URL: https://arxiv.org/abs/2106.09840

zbMATH Keywords

Berry-Esseen theorementrywise eigenvector analysisrandom dot product graphssignal-plus-noise matrix modelsymmetric noisy matrix completion

Mathematics Subject Classification ID

Multivariate analysis (62Hxx) Basic linear algebra (15Axx) Graph theory (05Cxx)

Cites Work

• Matrix completion from noisy entries
• Introduction to empirical processes and semiparametric inference
• Asymptotic Statistics
• Spectral clustering and the high-dimensional stochastic blockmodel
• Title not available (Why is that?)
• The Rotation of Eigenvectors by a Perturbation. III
• Spectral analysis of large dimensional random matrices
• Network cross-validation by edge sampling
• Matrix estimation by universal singular value thresholding
• Consistency of spectral clustering in stochastic block models
• Mixed membership stochastic blockmodels
• Title not available (Why is that?)
• Title not available (Why is that?)
• Community structure in social and biological networks
• Community detection and stochastic block models: recent developments
• On consistency and sparsity for principal components analysis in high dimensions
• Achieving optimal misclassification proportion in stochastic block models
• Random Dot Product Graph Models for Social Networks
• Exact matrix completion via convex optimization
• Nuclear-norm penalization and optimal rates for noisy low-rank matrix completion
• A generalized solution of the orthogonal Procrustes problem
• Tight Oracle Inequalities for Low-Rank Matrix Recovery From a Minimal Number of Noisy Random Measurements
• The Power of Convex Relaxation: Near-Optimal Matrix Completion
• A nonparametric two-sample hypothesis testing problem for random graphs
• Rate-optimal perturbation bounds for singular subspaces with applications to high-dimensional statistics
• Entrywise eigenvector analysis of random matrices with low expected rank
• Exact Recovery in the Stochastic Block Model
• Title not available (Why is that?)
• Perturbation bounds in connection with singular value decomposition
• Simple: Statistical Inference on Membership Profiles in Large Networks
• Asymptotic Theory of Eigenvectors for Random Matrices With Diverging Spikes
• Compressed sensing
• Random matrix theory in statistics: a review
• A Consistent Adjacency Spectral Embedding for Stochastic Blockmodel Graphs
• Semidefinite programming based preconditioning for more robust near-separable nonnegative matrix factorization
• A limit theorem for scaled eigenvectors of random dot product graphs
• Universally consistent vertex classification for latent positions graphs
• Perfect clustering for stochastic blockmodel graphs via adjacency spectral embedding
• Reconstruction of a low-rank matrix in the presence of Gaussian noise
• Tensor decompositions for learning latent variable models
• Recovering Low-Rank Matrices From Few Coefficients in Any Basis
• Low-rank matrix completion using alternating minimization
• Role of normalization in spectral clustering for stochastic blockmodels
• Efficient discovery of overlapping communities in massive networks
• The eigenvalues and eigenvectors of finite, low rank perturbations of large random matrices
• Minimax risk of matrix denoising by singular value thresholding
• Large sample covariance matrices and high-dimensional data analysis
• On the Monotonicity, Log-Concavity, and Tight Bounds of the Generalized Marcum and Nuttall $Q$-Functions
• Statistical inference on random dot product graphs: a survey
• Signal-plus-noise matrix models: eigenvector deviations and fluctuations
• The two-to-infinity norm and singular subspace geometry with applications to high-dimensional statistics
• Limit theorems for eigenvectors of the normalized Laplacian for random graphs
• Debiasing the Lasso: optimal sample size for Gaussian designs
• An \(\ell_{\infty}\) eigenvector perturbation bound and its application
• Unperturbed: spectral analysis beyond Davis-Kahan
• Near-optimal bounds for phase synchronization
• The Sup-norm Perturbation of HOSVD and Low Rank Tensor Denoising
• Estimating Mixed Memberships With Sharp Eigenvector Deviations
• A note on the orthogonal Procrustes problem and norm-dependent optimality
• Optimal Bayesian estimation for random dot product graphs
• Entrywise Estimation of Singular Vectors of Low-Rank Matrices With Heteroskedasticity and Dependence
• Efficient Estimation for Random Dot Product Graphs via a One-Step Procedure

This page was built for publication: Entrywise limit theorems for eigenvectors of signal-plus-noise matrix models with weak signals

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6178566)