Central limit theorem and self-normalized Cramér-type moderate deviation for Euler-Maruyama scheme

DOI10.3150/21-BEJ1372MaRDI QIDQ2137002zbMATH OpenOpenAlexWikidataFDO

Authors Jianya Lu, Yuzhen Tan, Lihu Xu

Publication date 16 May 2022

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2012.04328, https://projecteuclid.org/journals/bernoulli/volume-28/issue-2/Central-limit-theorem-and-self-normalized-Cram%c3%a9r-type-moderate-deviation/10.3150/21-BEJ1372.full

zbMATH Keywords

central limit theoremEuler-Maruyama schemeself-normalized Cramér-type moderate deviationStein's methodstochastic differential equation

Mathematics Subject Classification ID

Central limit and other weak theorems (60F05) Large deviations (60F10) Sums of independent random variables; random walks (60G50) Asymptotic distribution theory in statistics (62E20) Parametric hypothesis testing (62F03) Numerical solutions to stochastic differential and integral equations (65C30)

Abstract: We consider a stochastic differential equation and its Euler-Maruyama (EM) scheme, under some appropriate conditions, they both admit a unique invariant measure, denoted by $pi$ and $p{i}_{e}ta$ respectively ($eta$ is the step size of the EM scheme). We construct an empirical measure $P{i}_{e}ta$ of the EM scheme as a statistic of $p{i}_{e}ta$, and use Stein's method developed in citet{FSX19} to prove a central limit theorem of $P{i}_{e}ta$. The proof of the self-normalized Cram'er-type moderate deviation (SNCMD) is based on a standard decomposition on Markov chain, splitting $et{a}^{-1/2}(P{i}_{e}ta(.)-pi(.))$ into a martingale difference series sum $mcl{H}_{e}ta$ and a negligible remainder $mcl{R}_{e}ta$. We handle $mcl{H}_{e}ta$ by the time-change technique for martingale, while prove that $mcl{R}_{e}ta$ is exponentially negligible by concentration inequalities, which have their independent interest. Moreover, we show that SNCMD holds for $x=o(et{a}^{-1/6})$, which has the same order as that of the classical result in citet{shao1999cramer,JSW03}.

Recommendations

• Exact convergence rate of the Euler-Maruyama scheme, with application to sampling design
• Cramér type moderate deviation theorems for self-normalized processes
• Concentration inequalities for Euler schemes
• Recursive computation of the invariant distribution of a diffusion
• Self-normalized Cramér type moderate deviations for martingales

Cites work

• A Cramér type large deviation result for Student's \(t\)-statistic
• A refined Cramér-type moderate deviation for sums of local statistics
• Central limit theorem and self-normalized Cramér-type moderate deviation for Euler-Maruyama scheme
• Consistency and fluctuations for stochastic gradient Langevin dynamics
• Convergence rate of EM scheme for SDDEs
• Cramér type moderate deviation theorems for self-normalized processes
• Cramér type moderate deviations for self-normalized \(\psi \)-mixing sequences
• Cramér-type moderate deviation for Studentized compound Poisson sum
• Cramér-type moderate deviation theorems for nonnormal approximation
• Dependent central limit theorems and invariance principles
• Elliptic and Parabolic Second-Order PDEs with Growing Coefficients
• Exponential convergence of Langevin distributions and their discrete approximations
• From Stein identities to moderate deviations
• General state space Markov chains and MCMC algorithms
• High-dimensional Bayesian inference via the unadjusted Langevin algorithm
• scientific article; zbMATH DE number 2212010 (Why is no real title available?)
• Inequalities for a Pair of Processes Stopped at a Random Time
• Invariant Measures and Euler--Maruyama's Approximations of State-Dependent Regime-Switching Diffusions
• Lévy Processes and Stochastic Calculus
• Markov Chains and Stochastic Stability
• Markov chains for exploring posterior distributions. (With discussion)
• Moderate deviation principles for importance sampling estimators of risk measures
• Moderate deviations for interacting processes
• Moderate deviations-based importance sampling for stochastic recursive equations
• Multivariate approximations in Wasserstein distance by Stein's method and Bismut's formula
• New regularity of Kolmogorov equation and application on approximation of semi-linear SPDEs with Hölder continuous drifts
• Nonasymptotic bounds for sampling algorithms without log-concavity
• Nonasymptotic convergence analysis for the unadjusted Langevin algorithm
• Self-normalized Cramér type moderate deviations for martingales
• Self-normalized Cramér type moderate deviations for stationary sequences and applications
• Self-normalized Cramér-type large deviations for independent random variables.
• Self-normalized Cramér-type moderate deviations under dependence
• Stability of Markovian processes III: Foster–Lyapunov criteria for continuous-time processes
• Subgaussian concentration inequalities for geometrically ergodic Markov chains
• Theoretical Guarantees for Approximate Sampling from Smooth and Log-Concave Densities

Cited in

(10)

• Central limit theorem and self-normalized Cramér-type moderate deviation for Euler-Maruyama scheme
• CLT for approximating ergodic limit of SPDEs via a full discretization
• Cramér-type moderate deviations under local dependence
• Approximation of the invariant measure of stable SDEs by an Euler-Maruyama scheme
• Normalized and self-normalized Cramér-type moderate deviations for the Euler-Maruyama scheme for the SDE
• On numerical discretizations that preserve probabilistic limit behaviors for time-homogeneous Markov processes
• Gradient estimates for semigroups associated with stochastic differential equations driven by cylindrical Lévy processes
• Limit theorems for SDEs with irregular drifts
• Central limit theorem for temporal average of backward Euler-Maruyama method
• Existence and non-existence of the CLT for a family of SDEs driven by stable processes

This page was built for publication: Central limit theorem and self-normalized Cramér-type moderate deviation for Euler-Maruyama scheme

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2137002)