39
15

A uniform Berry--Esseen theorem on MM-estimators for geometrically ergodic Markov chains

Abstract

Let {Xn}n0\{X_n\}_{n\ge0} be a VV-geometrically ergodic Markov chain. Given some real-valued functional FF, define Mn(α):=n1k=1nF(α,Xk1,Xk)M_n(\alpha):=n^{-1}\sum_{k=1}^nF(\alpha,X_{k-1},X_k), αAR\alpha\in\mathcal{A}\subset \mathbb {R}. Consider an MM estimator α^n\hat{\alpha}_n, that is, a measurable function of the observations satisfying Mn(α^n)minαAMn(α)+cnM_n(\hat{\alpha}_n)\leq \min_{\alpha\in\mathcal{A}}M_n(\alpha)+c_n with {cn}n1\{c_n\}_{n\geq1} some sequence of real numbers going to zero. Under some standard regularity and moment assumptions, close to those of the i.i.d. case, the estimator α^n\hat{\alpha}_n satisfies a Berry--Esseen theorem uniformly with respect to the underlying probability distribution of the Markov chain.

View on arXiv
Comments on this paper