Non-Asymptotic Gaussian Estimates for the Recursive Approximation of the Invariant Measure of a Diffusion - Université d'Évry Access content directly
Journal Articles Annales de l'Institut Henri Poincaré (B) Probabilités et Statistiques Year : 2020

Non-Asymptotic Gaussian Estimates for the Recursive Approximation of the Invariant Measure of a Diffusion

Abstract

We obtain non-asymptotic Gaussian concentration bounds for the difference between the invariant measure ν of an ergodic Brownian diffusion process and the empirical distribution of an approximating scheme with decreasing time step along a suitable class of (smooth enough) test functions f such that f − ν(f) is a coboundary of the infinitesimal generator. We show that these bounds can still be improved when the (squared) Fröbenius norm of the diffusion coefficient lies in this class. We apply these bounds to design computable non-asymptotic confidence intervals for the approximating scheme. As a theoretical application, we finally derive non-asymptotic deviation bounds for the almost sure Central Limit Theorem.
Fichier principal
Vignette du fichier
ARTICLE_REV_IHP_230518_DEF_HAL.pdf (729.48 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01321645 , version 1 (26-05-2016)
hal-01321645 , version 2 (29-06-2016)
hal-01321645 , version 3 (10-07-2017)
hal-01321645 , version 4 (23-05-2018)

Identifiers

Cite

Igor Honoré, Stephane Menozzi, Gilles Pagès. Non-Asymptotic Gaussian Estimates for the Recursive Approximation of the Invariant Measure of a Diffusion. Annales de l'Institut Henri Poincaré (B) Probabilités et Statistiques, 2020, 3, pp.1559-1605. ⟨hal-01321645v4⟩
3796 View
607 Download

Altmetric

Share

Gmail Facebook X LinkedIn More