Gaussian approximation of suprema of empirical processes

Abstract

This paper develops a new direct approach to approximating suprema of general empirical processes by a sequence of suprema of Gaussian processes, without taking the route of approximating whole empirical processes in the sup-norm. We prove an abstract approximation theorem applicable to a wide variety of statistical problems, such as construction of uniform confidence bands for functions. Notably, the bound in the main approximation theorem is nonasymptotic and the theorem allows for functions that index the empirical process to be unbounded and have entropy divergent with the sample size. The proof of the approximation theorem builds on a new coupling inequality for maxima of sums of random vectors, the proof of which depends on an effective use of Stein's method for normal approximation, and some new empirical process techniques. We study applications of this approximation theorem to local and series empirical processes arising in nonparametric estimation via kernel and series methods, where the classes of functions change with the sample size and are non-Donsker. Importantly, our new technique is able to prove the Gaussian approximation for the supremum type statistics under weak regularity conditions, especially concerning the bandwidth and the number of series functions, in those examples.

This is the full version of the paper published in at http://dx.doi.org/10.1214/14-AOS1230 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (9)

• Gaussian approximations and multiplier bootstrap for maxima of sums of high-dimensional random vectors
• Gaussian approximation of suprema of empirical processes
• Confidence bands in density estimation
• Anti-concentration and honest, adaptive confidence bands
• Moment inequalities for functions of independent random variables
• Global uniform risk bounds for wavelet deconvolution estimators
• Uniform limit theorems for wavelet density estimators
• A uniform functional law of the logarithm for the local empirical process
• Revisiting two strong approximation results of Dudley and Philipp

Cited by in corpus (23)

• Uniform Post Selection Inference for LAD Regression and Other Z-estimation problems
• Gaussian approximation of suprema of empirical processes
• Nonparametric modal regression
• Anti-concentration and honest, adaptive confidence bands
• Moving Beyond Sub-Gaussianity in High-Dimensional Statistics: Applications in Covariance Estimation and Linear Regression
• Asymptotic theory for density ridges
• DoubleML -- An Object-Oriented Implementation of Double Machine Learning in R
• Empirical Process Results for Exchangeable Arrays
• Are Discoveries Spurious? Distributions of Maximum Spurious Correlations and Their Applications
• Learning with Density Matrices and Random Features
• A Simple Adjustment for Bandwidth Snooping
• Honest Confidence Sets in Nonparametric IV Regression and Other Ill-Posed Models
• Notes on the dimension dependence in high-dimensional central limit theorems for hyperrectangles
• Inference for First-Price Auctions with Guerre, Perrigne, and Vuong's Estimator
• Unsupervised non-parametric change point detection in quasi-periodic signals
• Monotonicity-Constrained Nonparametric Estimation and Inference for First-Price Auctions
• Stratified incomplete local simplex tests for curvature of nonparametric multiple regression
• Bump hunting through density curvature features
• Trimmed sample means for robust uniform mean estimation and regression
• Inference on Individual Treatment Effects in Nonseparable Triangular Models
• Some New Asymptotic Theory for Least Squares Series: Pointwise and Uniform Results
• Bias correction for quantile regression estimators
• Continuous difference-in-differences with double/debiased machine learning