Entrywise Estimation of Singular Vectors of Low-Rank Matrices with Heteroskedasticity and Dependence

Abstract

We propose an estimator for the singular vectors of high-dimensional low-rank matrices corrupted by additive subgaussian noise, where the noise matrix is allowed to have dependence within rows and heteroskedasticity between them. We prove finite-sample ${\ell }_{2,\infty }$ bounds and a Berry-Esseen theorem for the individual entries of the estimator, and we apply these results to high-dimensional mixture models. Our Berry-Esseen theorem clearly shows the geometric relationship between the signal matrix, the covariance structure of the noise, and the distribution of the errors in the singular vector estimation task. These results are illustrated in numerical simulations. Unlike previous results of this type, which rely on assumptions of gaussianity or independence between the entries of the additive noise, handling the dependence between entries in the proofs of these results requires careful leave-one-out analysis and conditioning arguments. Our results depend only on the signal-to-noise ratio, the sample size, and the spectral properties of the signal matrix.

References in corpus (10)

• Spectral Methods for Data Science: A Statistical Perspective
• Normal approximation and concentration of spectral projectors of sample covariance
• The geometry of kernelized spectral clustering
• Inference for Heteroskedastic PCA with Missing Data
• Nonparametric two-sample hypothesis testing for low-rank random graphs of differing sizes
• A Schatten-$q$ Low-rank Matrix Perturbation Analysis via Perturbation Projection Error Bound
• Minimax Estimation of Linear Functions of Eigenvectors in the Face of Small Eigen-Gaps
• Hypothesis Testing for Equality of Latent Positions in Random Graphs
• Euclidean Representation of Low-Rank Matrices and Its Statistical Applications
• An ${\ell }_p$ theory of PCA and spectral clustering

Cited by in corpus (2)

• Inference for Heteroskedastic PCA with Missing Data
• Matrix Denoising with Partial Noise Statistics: Optimal Singular Value Shrinkage of Spiked F-Matrices