Central limit theorem and near classical Berry-Esseen rate for self normalized sums in high dimensions

DOI10.3150/23-BEJ1597zbMATH Open1530.60023arXiv2012.03758OpenAlexW4388513583MaRDI QIDQ6178560FDOQ6178560

Authors: Debraj Das 

Publication date: 16 January 2024

Published in: Bernoulli (Search for Journal in Brave)

Abstract: In this article, we are interested in the high dimensional normal approximation of $T_n=Big(su{m}_{i=1}^n{X}_{i1}/Big(sqrtsu{m}_{i=1}^n{X}_{i1}^{2}Big),dots,$ $su{m}_{i=1}^n{X}_{ip}/Big(sqrtsu{m}_{i=1}^n{X}_{ip}^{2}Big)Big)$ in $mathcal{R}^p$ uniformly over the class of hyper-rectangles $mathcal{A}^{re}=pro{d}_{j=1}^p[a_j,b_j]capmathcalR:-inftyleq{a}_{j}leq{b}_{j}leqinfty,j=1,ldots,p$, where $X_1,dots,X_n$ are non-degenerate independent $p-$dimensional random vectors. We assume that the components of $X_i$ are independent and identically distributed (iid) and investigate the optimal cut-off rate of $logp$ in the uniform central limit theorem (UCLT) for $T_n$ over $mathcal{A}^{re}$. The aim is to reduce the exponential moment conditions, generally assumed for exponential growth of the dimension with respect to the sample size in high dimensional CLT, to some polynomial moment conditions. Indeed, we establish that only the existence of some polynomial moment of order $in[2,4]$ is sufficient for exponential growth of $p$. However the rate of growth of $logp$ can not further be improved from $$ as a power of $n$ even if $X_{ij}$'s are iid across $(i,j)$ and $X_{11}$ is bounded. We also establish near$-n^{-kappa/2}$ Berry-Esseen rate for $T_n$ in high dimension under the existence of $(2+kappa)$th absolute moments of $X_{ij}$ for $0<kappaleq1$. When $kappa=1$, the obtained Berry-Esseen rate is also shown to be optimal. As an application, we find respective versions for component-wise Student's t-statistic, which may be useful in high dimensional statistical inference.

Full work available at URL: https://arxiv.org/abs/2012.03758

Recommendations

• Central limit theorem in high dimensions: the optimal bound on dimension growth rate
• High-dimensional CLT: improvements, non-uniform extensions and large deviations
• Notes on the dimension dependence in high-dimensional central limit theorems for hyperrectangles
• An exponential nonuniform Berry-Esseen bound for self-normalized sums
• Central limit theorem and bootstrap approximation in high dimensions: near \(1/\sqrt{n}\) rates via implicit smoothing

zbMATH Keywords

Berry-Esseen theoremself-normalized sumStudent \(t\)-statisticuniform central limit theorem (UCLT)

Mathematics Subject Classification ID

Central limit and other weak theorems (60F05)

Cites Work

• Central limit theorems and bootstrap in high dimensions
• Gaussian approximations and multiplier bootstrap for maxima of sums of high-dimensional random vectors
• Self-Normalized Processes
• Empirical and multiplier bootstraps for suprema of empirical processes of increasing complexity, and related Gaussian couplings
• An Inequality for Mill's Ratio
• Two moments suffice for Poisson approximations: The Chen-Stein method
• An exponential nonuniform Berry-Esseen bound for self-normalized sums
• A Remark on Stirling's Formula
• Title not available (Why is that?)
• The Influence of the Maximum Term in the Addition of Independent Random Variables
• When is the Student \(t\)-statistic asymptotically standard normal?
• Donsker's theorem for self-normalized partial sums processes
• The Berry-Esseen bound for Student's statistic
• A Berry-Esséen bound for Student's statistic in the non-i. i. d. case
• Limit distributions of self-normalized sums
• Self-normalized limit theorems: a survey
• Student's t-Test Under Symmetry Conditions
• Self-normalized Cramér-type large deviations for independent random variables.
• On the self-normalized Cramér-type large deviation
• Title not available (Why is that?)
• Further refinement of self-normalized Cramér-type moderate deviations
• Limit distributions of Studentized means.
• A Cramér type large deviation result for Student's \(t\)-statistic
• Self-normalized large deviations
• Refined self-normalized large deviations for independent random variables
• Limit theorems for self-normalized large deviation
• On the effect of random norming on the rate of convergence in the central limit theorem
• On Bounds for Moderate Deviations for Student's Statistic
• The Berry-Esseen bound for self-normalized sums
• Title not available (Why is that?)
• Notes on the dimension dependence in high-dimensional central limit theorems for hyperrectangles
• High-dimensional central limit theorems by Stein's method
• Quantile coupling inequalities and their applications
• Central limit theorem and bootstrap approximation in high dimensions: near \(1/\sqrt{n}\) rates via implicit smoothing
• Central limit theorem in high dimensions: the optimal bound on dimension growth rate

This page was built for publication: Central limit theorem and near classical Berry-Esseen rate for self normalized sums in high dimensions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6178560)