Beyond Gaussian Approximation: Bootstrap for Maxima of Sums of Independent Random Vectors

TL;DR

This paper advances bootstrap methods for high-dimensional maxima of sums of independent vectors, reducing sample size requirements and developing new theoretical tools beyond Gaussian approximation.

Contribution

It introduces bootstrap validity under weaker sample size conditions and develops novel comparison and anti-concentration theorems for high-dimensional maxima.

Findings

Bootstrap methods are consistent for n  ( p)^5

New comparison theorems improve high-dimensional probability bounds

Anti-concentration results support bootstrap validity without Gaussian approximation

Abstract

The Bonferroni adjustment, or the union bound, is commonly used to study rate optimality properties of statistical methods in high-dimensional problems. However, in practice, the Bonferroni adjustment is overly conservative. The extreme value theory has been proven to provide more accurate multiplicity adjustments in a number of settings, but only on ad hoc basis. Recently, Gaussian approximation has been used to justify bootstrap adjustments in large scale simultaneous inference in some general settings when $n \gg (\log p)^7$, where $p$ is the multiplicity of the inference problem and $n$ is the sample size. The thrust of this theory is the validity of the Gaussian approximation for maxima of sums of independent random vectors in high-dimension. In this paper, we reduce the sample size requirement to $n \gg (\log p)^5$ for the consistency of the empirical bootstrap and the multiplier/wild bootstrap in the Kolmogorov-Smirnov distance, possibly in the regime where the Gaussian approximation is not available. New comparison and anti-concentration theorems, which are of considerable interest in and of themselves, are developed as existing ones interweaved with Gaussian approximation are no longer applicable.