The Annals of Probability

Universal Donsker Classes and Metric Entropy

R. M. Dudley

Full-text: Open access

Abstract

Let $(X, \mathscr{A})$ be a measurable space and $\mathscr{F}$ a class of measurable functions on $X. \mathscr{F}$ is called a universal Donsker class if for every probability measure $P$ on $\mathscr{A}$, the centered and normalized empirical measures $n^{1/2}(P_n - P)$ converge in law, with respect to uniform convergence over $\mathscr{F}$, to the limiting "Brownian bridge" process $G_P$. Then up to additive constants, $\mathscr{F}$ must be uniformly bounded. Several nonequivalent conditions are shown to imply the universal Donsker property. Some are connected with the Vapnik-Cervonenkis combinatorial condition on classes of sets, others with metric entropy. The implications between the various conditions are considered. Bounds are given for the metric entropy of convex hulls in Hilbert space.

Article information

Source
Ann. Probab., Volume 15, Number 4 (1987), 1306-1326.

Dates
First available in Project Euclid: 19 April 2007

Permanent link to this document
https://projecteuclid.org/euclid.aop/1176991978

Digital Object Identifier
doi:10.1214/aop/1176991978

Mathematical Reviews number (MathSciNet)
MR905333

Zentralblatt MATH identifier
0631.60004

JSTOR
links.jstor.org

Subjects
Primary: 60F17: Functional limit theorems; invariance principles
Secondary: 60F05: Central limit and other weak theorems 60G17: Sample path properties 60G20: Generalized stochastic processes

Keywords
Central limit theorems Vapnik-Cervonenkis classes

Citation

Dudley, R. M. Universal Donsker Classes and Metric Entropy. Ann. Probab. 15 (1987), no. 4, 1306--1326. doi:10.1214/aop/1176991978. https://projecteuclid.org/euclid.aop/1176991978


Export citation