Irregular sets and Central Limit Theorems for dependent triangular arrays

Original · EN

In previous papers, we studied the asymptotic behaviour of Sₙ(A,X)=(2N+1)⁻ᵈ/²∑ₙ ∈ ₐₙ Xₙ, where X is a centered, stationary and weakly dependent random field, and Aₙ=A ∩ [-N,N]ᵈ, A ⊂ Zᵈ. This leads to the definition of asymptotically measurable sets, which enjoy the property that Sₙ(A;X) has a Gaussian weak limit for any X belonging to a certain class. Here we extend this type of results to the case of weakly dependent triangular arrays and present an application of this technique to regression models. Indeed, we prove that CLT and related results hold for Xₙⁿ=φ(ξₙⁿ,Yₙⁿ), n ∈ Zᵈ, where φ satisfies certain regularity conditions, ξ and Y are independent random fields, ξ is weakly dependent and Y satisfies some Strong Law of Large Numbers.