High-dimensional Central Limit Theorems by Stein's Method

Abstract

We obtain explicit error bounds for the $d$-dimensional normal approximation on hyperrectangles for a random vector that has a Stein kernel, or admits an exchangeable pair coupling, or is a non-linear statistic of independent random variables or a sum of $n$ locally dependent random vectors. We assume the approximating normal distribution has a non-singular covariance matrix. The error bounds vanish even when the dimension $d$ is much larger than the sample size $n$. We prove our main results using the approach of Götze (1991) in Stein's method, together with modifications of an estimate of Anderson, Hall and Titterington (1998) and a smoothing inequality of Bhattacharya and Rao (1976). For sums of $n$ independent and identically distributed isotropic random vectors having a log-concave density, we obtain an error bound that is optimal up to a $\log n$ factor. We also discuss an application to multiple Wiener-Itô integrals.

arXiv comment31 pages. Major update. Extension to the non-singular case

References in corpus (8)

• A new method of normal approximation
• Multivariate normal approximation with Stein's method of exchangeable pairs under a general linearity condition
• Chaos, concentration, and multiple valleys
• Stein kernels and moment maps
• Approximating high-dimensional infinite-order $U$-statistics: statistical and computational guarantees
• Explicit rates of approximation in the CLT for quadratic forms
• Notes on the dimension dependence in high-dimensional central limit theorems for hyperrectangles
• Multivariate normal approximation on the Wiener space: new bounds in the convex distance

Cited by in corpus (1)

• Strong Gaussian Approximation for the Sum of Random Vectors