Concentration inequalities and moment bounds for sample covariance operators

DOI10.3150/15-BEJ730MaRDI QIDQ502859zbMATH OpenOpenAlexFDO

Authors K. Lounici, Vladimir Koltchinskii

Publication date 11 January 2017

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1405.2468, https://projecteuclid.org/euclid.bj/1475001350

zbMATH Keywords

concentration inequalitiesmoment boundsGaussiansample covariance

Mathematics Subject Classification ID

Gaussian processes (60G15) Probability theory on linear topological spaces (60B11) Inequalities; stochastic orderings (60E15)

Abstract: Let $X,X_1,dots,X_n,dots$ be i.i.d. centered Gaussian random variables in a separable Banach space $E$ with covariance operator $Sigma:$ ${\displaystyle }$ Sigma:E^{ast}mapsto E, Sigma u = {mathbb E}langle X,u angle, uin E^{ast}. ${\displaystyle }$ The sample covariance operator $hatSigma:E^{ast}mapstoE$ is defined as ${\displaystyle }$ hat Sigma u := n^{-1}sum_{j=1}^n langle X_j,u angle X_j, uin E^{ast}. ${\displaystyle }$ The goal of the paper is to obtain concentration inequalities and expectation bounds for the operator norm $|hatSigma-Sigma|$ of the deviation of the sample covariance operator from the true covariance operator. In particular, it is shown that ${\displaystyle }$ {mathbb E}|hat Sigma-Sigma|asymp |Sigma|�iggl(sqrt{frac{{�f r}(Sigma)}{n}}�igvee frac{{�f r}(Sigma)}{n}�iggr), ${\displaystyle }$ where ${\displaystyle }$ {�f r}(Sigma):=frac{Bigl({mathbb E}|X|Bigr)^2}{|Sigma|}. ${\displaystyle }$ Moreover, under the assumption that $$ it is proved that, for all $tgeq1,$ with probability at least $1-e^{-t}$ �egin{align*} Bigl||hatSigma - Sigma|-{mathbb E}|hatSigma - Sigma|Bigr| lesssim |Sigma|�iggl(sqrt{frac{t}{n}}�igvee frac{t}{n}�iggr). end{align*}

Recommendations

• Normal approximation and concentration of spectral projectors of sample covariance
• New concentration inequalities for suprema of empirical processes
• Relative perturbation bounds with applications to empirical covariance operators
• Asymptotics and concentration bounds for bilinear forms of spectral projectors of sample covariance
• Concentration inequalities for random tensors

Cited in

(77)

• A shrinkage principle for heavy-tailed data: high-dimensional robust low-rank matrix recovery
• High-resolution signal recovery via generalized sampling and functional principal component analysis
• Robust modifications of U-statistics and applications to covariance estimation problems
• New challenges in covariance estimation: multiple structures and coarse quantization
• Covariance estimation under one-bit quantization
• Finite-sample analysis of \(M\)-estimators using self-concordance
• Contiguity under high-dimensional Gaussianity with applications to covariance testing
• Bootstrapping the operator norm in high dimensions: error estimation for covariance matrices and sketching
• Noise covariance estimation in multi-task high-dimensional linear models
• Time-uniform, nonparametric, nonasymptotic confidence sequences
• Uniform Hanson-Wright type concentration inequalities for unbounded entries via the entropy method
• A note on the prediction error of principal component regression in high dimensions
• Relative perturbation bounds with applications to empirical covariance operators
• The Sup-norm Perturbation of HOSVD and Low Rank Tensor Denoising
• Robustifying Markowitz
• Robust covariance estimation under \(L_4\)-\(L_2\) norm equivalence
• Estimating covariance and precision matrices along subspaces
• Fast randomized numerical rank estimation for numerically low-rank matrices
• Sub-Gaussian estimators of the mean of a random matrix with heavy-tailed entries
• On consistency and sparsity for high-dimensional functional time series with application to autoregressions
• All-in-one robust estimator of the Gaussian mean
• Composite Index Construction with Expert Opinion
• Large-dimensional central limit theorem with fourth-moment error bounds on convex sets and balls
• On the non-asymptotic concentration of heteroskedastic Wishart-type matrix
• Construction and Monte Carlo estimation of wavelet frames generated by a reproducing kernel
• Bootstrapping max statistics in high dimensions: near-parametric rates under weak variance decay and application to functional and multinomial data
• Cross-validation with confidence
• CDPA: common and distinctive pattern analysis between high-dimensional datasets
• Tangent space and dimension estimation with the Wasserstein distance
• Covariance estimation under missing observations and \(L_4 - L_2\) moment equivalence
• On cumulative slicing estimation for high dimensional data
• The two-to-infinity norm and singular subspace geometry with applications to high-dimensional statistics
• Model Reduction for Nonlinear Systems by Balanced Truncation of State and Gradient Covariance
• Convexification with bounded gap for randomly projected quadratic optimization
• Wald Statistics in high-dimensional PCA
• SONIC: social network analysis with influencers and communities
• Hanson-Wright inequality in Banach spaces
• Efficient estimation of smooth functionals in Gaussian shift models
• An elementary analysis of ridge regression with random design
• Quantitative limit theorems and bootstrap approximations for empirical spectral projectors
• Robust high-dimensional factor models with applications to statistical machine learning
• Approximating \(L_p\) unit balls via random sampling
• Dimension-free bounds for sums of independent matrices and simple tensors via the variational principle
• Asymptotically efficient estimation of smooth functionals of covariance operators
• Minimax rates in sparse, high-dimensional change point detection
• Finite impulse response models: a non-asymptotic analysis of the least squares estimator
• Optimal estimation of Schatten norms of a rectangular matrix
• New asymptotic results in principal component analysis
• Efficient estimation of linear functionals of principal components
• Bayesian inference for spectral projectors of the covariance matrix
• Time-uniform Chernoff bounds via nonnegative supermartingales
• Convergence rate of Krasulina estimator
• Distributed estimation of principal eigenspaces
• Mean estimation in high dimension
• Estimation of smooth functionals in normal models: bias reduction and asymptotic efficiency
• Dimension-free bounds for sums of dependent matrices and operators with heavy-tailed distributions
• Non-zero constraints in elliptic PDE with random boundary values and applications to hybrid inverse problems
• On convergence rates of adaptive ensemble Kalman inversion for linear ill-posed problems
• Bootstrap confidence sets for spectral projectors of sample covariance
• Perturbation bounds for eigenspaces under a relative gap condition
• Model assisted variable clustering: minimax-optimal recovery and algorithms
• Improved covariance estimation: optimal robustness and sub-Gaussian guarantees under heavy tails
• Deep learning: a statistical viewpoint
• scientific article; zbMATH DE number 7625199 (Why is no real title available?)
• Nonasymptotic upper bounds for the reconstruction error of PCA
• Dimensionality Reduction, Regularization, and Generalization in Overparameterized Regressions
• Detecting approximate replicate components of a high-dimensional random vector with latent structure
• Moment bounds for large autocovariance matrices under dependence
• Exact minimax risk for linear least squares, and the lower tail of sample covariance matrices
• Matrix concentration inequalities and free probability
• Confidence sets for spectral projectors of covariance matrices
• Learning Gaussian graphical models with latent confounders
• MSE bounds for estimators of matrix functions
• Nonasymptotic one- and two-sample tests in high dimension with unknown covariance structure
• Optimal multiple change-point detection for high-dimensional data
• An \({\ell_p}\) theory of PCA and spectral clustering
• Normal approximation and concentration of spectral projectors of sample covariance

This page was built for publication: Concentration inequalities and moment bounds for sample covariance operators

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q502859)