Approximation of stable law in Wasserstein-1 distance by Stein's method

DOI10.1214/18-AAP1424zbMATH Open1448.60062OpenAlexW2784566409WikidataQ128891835 ScholiaQ128891835MaRDI QIDQ670747FDOQ670747

Authors Lihu Xu 

Publication date 20 March 2019

Published in The Annals of Applied Probability (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1709.00805

zbMATH Keywords

Stein's method\(\alpha\)-stable processes\(L^1\) discrepancynormal domain of attraction of stable lawstable approximationWasserstein-1 distance (\(W_1\) distance)

Mathematics Subject Classification ID

Infinitely divisible distributions; stable distributions (60E07) Central limit and other weak theorems (60F05) Sums of independent random variables; random walks (60G50) Stable stochastic processes (60G52)

Abstract: Let $ninmathbbN$, let $zet{a}_{n,1},\mathrm{...},zet{a}_{n,n}$ be a sequence of independent random variables with $mathbbEzet{a}_{n,i}=0$ and $mathbbE|zet{a}_{n,i}|<infty$ for each $i$, and let $mu$ be an $alpha$-stable distribution having characteristic function $e^{-|lambda{|}^{alpha}}$ with $alphain(1,2)$. Denote $S_n=zet{a}_{n,1}+\mathrm{...}+zet{a}_{n,n}$ and its distribution by $mathcalL(S_n)$, we bound the Wasserstein distance of $mathcalL(S_n)$ and $mu$ essentially by an $L^1$ discrepancy between two kernels, this bound can be interpreted as a generalization of the Stein discrepancy (in $L^2$ sense) introduced by Ledoux, Nourdin and Peccati. More precisely, we prove the following inequality: �egin{equation} �egin{split} d_Wleft(mathcal L (S_n), mu ight) le C left[sum_{i=1}^nint_{-N}^N left|frac{mathcal K_alpha(t,N)}n -frac{ K_i(t,N)}{alpha} ight| d t + mathcal R_{N,n} ight], end{split} end{equation} where $d_W$ is the Wasserstein distance of probability measures, $mathcal{K}_{a}lpha(t,N)$ is the kernel of a decomposition of the fractional Laplacian $Delt{a}^{fracalpha2}$, $K_i(t,N)$ is a kernel introduced by Chen, Goldstein and Shao with a truncation which can be interpreted as an $L^1$ Stein kernel, and $mathcal{R}_{N,n}$ is a small remainder. The integral term ${\displaystyle }$sum_{i=1}^nint_{-N}^N left|frac{mathcal K_alpha(t,N)}n -frac{ K_i(t,N)}{alpha} ight| d t${\displaystyle }$ can be interpreted as an $L^1$ Stein discrepancy. As an application, we prove a general theorem of stable law convergence rate when $zet{a}_{n,i}$ are i.i.d. and the distribution falls in the normal domain of attraction of $mu$. We also study four examples with comparing our convergence rates and those known for these examples, among which the distribution in the second example is not in the normal domain of attraction of $mu$.

Recommendations

• Approximation de la L1-distance de Wasserstein
• Stein's method and Poisson process approximation for a class of Wasserstein metrics
• Stein's method for normal approximation in Wasserstein distances with application to the multivariate central limit theorem
• The Wasserstein distance and approximation theorems
• Approximation for the Wasserstein distance
• Multivariate approximations in Wasserstein distance by Stein's method and Bismut's formula
• Stein factors for variance-gamma approximation in the Wasserstein and Kolmogorov distances
• Approximation rate in Wasserstein distance of probability measures on the real line by deterministic empirical measures
• On Stein's factors for Poisson approximation in Wasserstein distance
• A weighted approximation approach to the study of the empirical Wasserstein distance

Cites Work

• scientific article; zbMATH DE number 3161601 (Why is no real title available?)
• scientific article; zbMATH DE number 3438144 (Why is no real title available?)
• scientific article; zbMATH DE number 204994 (Why is no real title available?)
• scientific article; zbMATH DE number 1889798 (Why is no real title available?)
• scientific article; zbMATH DE number 2212010 (Why is no real title available?)
• An extended Stein-type covariance identity for the Pearson family with applications to lower variance bounds
• Central limit theorem and convergence to stable laws in Mallows distance
• Chance and Stability
• Chi-square approximation by Stein's method with application to Pearson's statistic
• Degree asymptotics with rates for preferential attachment random graphs
• Discretized normal approximation by Stein's method
• Entropy and the fourth moment phenomenon
• Ergodicity for time-changed symmetric stable processes
• Fluctuations of eigenvalues and second order Poincaré inequalities
• General theorems on rates of convergence in distribution of random variables II. Applications to the stable limit laws and weak law of large numbers
• Heat kernel estimates for stable-like processes on \(d\)-sets.
• Heat kernels and analyticity of non-symmetric jump diffusion semigroups
• Invariant measures and symmetry property of Lévy type operators
• Limit Theorems for Spectra of Random Matrices with Martingale Structure
• Multivariate normal approximation with Stein's method of exchangeable pairs under a general linearity condition
• Nonnormal approximation by Stein's method of exchangeable pairs with application to the Curie-Weiss model
• Normal Approximation by Stein’s Method
• Normal approximations with Malliavin calculus. From Stein's method to universality
• On Stein's method and perturbations
• On Stein's method for infinitely divisible laws with finite first moment
• On the Rate of Convergence to a Stable Law
• On two approaches to approximation of multidimensional stable laws
• Poisson approximation for dependent trials
• Probability. Theory and examples.
• Rate of convergence to a stable law
• Some remarks on the rate of convergence to stable laws
• Stable convergence and stable limit theorems
• Stein factors for negative binomial approximation in Wasserstein distance
• Stein's method and the zero bias transformation with application to simple random sampling
• Stein's method for comparison of univariate distributions
• Stein's method for concentration inequalities
• Stein's method for dependent random variables occuring in statistical mechanics
• Stein's method for invariant measures of diffusions via Malliavin calculus
• Stein's method for steady-state diffusion approximations of \(\mathrm{M}/\mathrm{Ph}/n+\mathrm{M}\) systems
• Stein's method of exchangeable pairs for the Beta distribution and generalizations
• Stein's method on Wiener chaos
• Stein's method, logarithmic Sobolev and transport inequalities
• Symmetric Stable Laws and Stable-Like Jump-Diffusions
• Ten equivalent definitions of the fractional Laplace operator
• The Accuracy of Infinitely Divisible Approximations to Sums of Independent Variables with Application to Stable Laws
• Two-sided bounds on the rate of convergence to a stable law

Cited In (21)

• A probability approximation framework: Markov process approach
• On Stein's method for multivariate self-decomposable laws
• Approximation of the invariant measure of stable SDEs by an Euler-Maruyama scheme
• Cramér-type moderate deviations under local dependence
• Non-integrable stable approximation by Stein's method
• Gaussian random field approximation via Stein's method with applications to wide random neural networks
• On algebraic Stein operators for Gaussian polynomials
• A unified approach to Stein's method for stable distributions
• On Stein factors for Laplace approximation and their application to random sums
• Stein factors for variance-gamma approximation in the Wasserstein and Kolmogorov distances
• On Brascamp-Lieb and Poincaré type inequalities for generalized tempered stable distribution
• Stein's method for asymmetric \(\alpha \)-stable distributions, with application to the stable CLT
• Limits of invariant measures of stochastic Burgers equations driven by two kinds of \(\alpha\)-stable processes
• A kernel bound for non-symmetric stable distribution and its applications
• Approximation to stable law by the Lindeberg principle
• Multivariate stable approximation by Stein's method
• Computing the exponent of a Lebesgue space
• A Stein characterisation of the distribution of the product of correlated normal random variables
• Stable central limit theorem in total variation distance
• First-order covariance inequalities via Stein's method
• A generalized Catoni's M-estimator under finite \(\alpha\)-th moment assumption with \(\alpha \in (1,2)\)

This page was built for publication: Approximation of stable law in Wasserstein-1 distance by Stein's method

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q670747)