Moderate deviations principle for empirical covariance from a unit root

MaRDI QIDQ6234442zbMATH OpenFDO

Authors Yu Miao, Yanling Wang, Guangyu Yang

Publication date 17 July 2012

Mathematics Subject Classification ID

Large deviations (60F10) Stationary stochastic processes (60G10) Linear regression; mixed models (62J05)

Abstract: In the present paper, we consider the linear autoregressive model in $r$, ${\displaystyle }$ X_{k,n}= heta_n X_{k,n-1}+xi_k, k=0,1,...,n, nge 1${\displaystyle }$ where $het{a}_{n}in[0,1)$ is unknown, $(x{i}_k{)}_{kinzz}$ is a sequence of centered i.i.d. r.v. valued in $r$ representing the noise. When $het{a}_{n}o1$, the moderate deviations principle for empirical covariance is discussed and as statistical applications we provide the moderate deviation estimates of the least square and the Yule-Walker estimators of the parameter $het{a}_n$.

This page was built for publication: Moderate deviations principle for empirical covariance from a unit root

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6234442)