Abstract:The central limit theorem is one of the most fundamental results in probability and has been successfully extended to locally dependent data and strongly-mixing random fields. In this paper, we establish its rate of convergence for transport distances, namely for arbitrary $p\ge1$ we obtain an upper bound for the Wasserstein-$p$ distance for locally dependent random variables and strongly mixing stationary random fields. Our proofs adapt the Stein dependency neighborhood method to the Wasserstein-$p$ distance and as a by-product we establish high-order local expansions of the Stein equation for dependent random variables. Finally, we demonstrate how our results can be used to obtain tail bounds that are asymptotically tight, and decrease polynomially fast, for the empirical average of weakly dependent random variables.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the convergence rate of the central limit theorem under weakly dependent data. Specifically, the goal of the paper is to establish the upper bound of the Wasserstein - p distance between the standardized sum and the normal distribution for locally dependent random variables and strongly mixing random fields. This goal is achieved by extending the Stein method to handle the Wasserstein distance of any order $p \geq 1$ and providing new tools to study the deviation behavior of dependent random variables. ### Background of the Paper The central limit theorem is one of the most fundamental results in probability theory. It was originally proposed for independent and identically distributed random variables. Later, it was extended to locally dependent data and strongly mixing random fields. However, for the general Wasserstein - p distance ($p \geq 1$), there are no known similar upper bounds under dependent data. This paper fills this gap. ### Main Contributions 1. **Establishing the Upper Bound of the Wasserstein - p Distance**: - For locally dependent random variables and strongly mixing stationary random fields, the paper establishes the upper bound of the Wasserstein - p distance. - Specifically, for a d - dimensional stationary random field, if the mixing coefficient $\alpha_\ell = O(\ell^{-\beta})$ and $\beta > d(p + 1)$, then: \[ W_p(L(W_n), N(0, 1)) = O\left(\frac{1}{\sqrt{|I_n|}}\right) \] - If the mixing coefficient $\alpha_\ell=\ell^{-\beta}$ and $\beta\in\left(d(p + 1)/2, d(p + 1)\right]$, then: \[ W_p(L(W_n), N(0, 1)) = O\left(\frac{1}{|I_n|^\gamma}\right) \] where $\gamma < 1/2$ is an explicit constant depending on $\beta$. 2. **Extending the Stein Method**: - A new way is proposed to adapt the Stein method to obtain the upper bound of the general Wasserstein - p distance. This is the first successful technique for dealing with a large class of dependent random variables. - This method is applicable to any real number $p \geq 1$ and can be extended to mixing random fields. 3. **Application to Tail Bounds**: - The paper shows how to use the Wasserstein - p distance ($p > 1$) to obtain tail bounds. Specifically, by choosing appropriate parameters, asymptotically tight tail bounds can be obtained. These bounds decrease polynomially fast with $t$ and are valid for mixing sequences. ### Application Examples - **m - Dependent Random Fields**: - For m - dependent random fields, if certain moment conditions and non - degeneracy conditions are satisfied, then: \[ W_p(L(W_n), N(0, 1)) = O\left(\frac{1}{\sqrt{|T_n|}}\right) \] - **U - Statistics**: - For U - statistics induced by symmetric functions, if the mean is zero, moment conditions and non - degeneracy conditions are satisfied, then: \[ W_p(L(W_n), N(0, 1)) = O\left(\frac{1}{\sqrt{n}}\right) \] ### Conclusion By extending the Stein method, this paper successfully establishes the upper bound of the convergence rate in the Wasserstein - p distance for locally dependent random variables and strongly mixing random fields. These results not only fill the theoretical gap but also provide useful tools in practical applications, especially in fields such as statistical inference and machine learning.

Wasserstein-p Bounds in the Central Limit Theorem Under Weak Dependence

Wasserstein-p Bounds in the Central Limit Theorem Under Local Dependence

Approximation of Sums of Locally Dependent Random Variables via Perturbation of Stein Operator

Limit theorems for kernel density estimators under dependent samples

Wasserstein-p Bounds via Cumulant-Based Edgeworth Expansion for $α$-Mixing Random Fields

The Functional Central Limit Theorem for Strong Near-Epoch Dependent Random Variables *

An Almost Sure Central Limit Theorem for Self-Normalized Partial Sums of Weakly Dependent Random Variables

Rates of convergence in the central limit theorem for Banach valued dependent variables

Wasserstein convergence rates of increasingly concentrating probability measures

Improved rates of convergence for the multivariate Central Limit Theorem in Wasserstein distance

Strong Limit Theorems for Extended Independent Random Variables and Extended Negatively Dependent Random Variables under Sub-Linear Expectations

Central limit theorems for stationary random fields under weak dependence with application to ambit and mixed moving average fields

Wasserstein and total variation distance between marginals of Lévy processes

Central limit theorems for high dimensional dependent data

Wasserstein convergence for conditional empirical measures of subordinated Dirichlet diffusions on Riemannian manifolds

Wasserstein-$1$ distance and nonuniform Berry-Esseen bound for a supercritical branching process in a random environment

Central limit theorem for the Sliced 1-Wasserstein distance and the max-Sliced 1-Wasserstein distance

Exponential convergence in Wasserstein metric for distribution dependent SDEs

On the Strong Convergence of Weighted Sums of Widely Dependent Random Variables

On Wasserstein-1 Distance in the Central Limit Theorem for Elephant Random Walk

Hoeffding's inequality for sums of weakly dependent random variables