A. Jach, T. McElroy

We provide a self-normalization for the sample acvs and acs of a linear, long-memory time series with innovations that have either finite 4th moment or are heavy-tailed with tail index $2<\alpha<4$. In the asymptotic distribution of the sample acv there are three rates of convergence that depend on the interplay between the memory parameter d and alpha, and which consequently lead to three different limit distributions; for the sample ac the limit distribution only depends on d. We introduce a self-normalized sample acv statistic, which is computable without knowledge of alpha or d, and which converges to a nondegenerate distribution. We also treat self-normalization of the acs. The sampling distributions can then be approximated nonparametrically by subsampling, as the asymptotic distribution is still parameter-dependent. The subsampling-based confidence intervals for the process acvs and acs are shown to have satisfactory empirical coverage rates in a simulation study.

Palabras clave: linear time series, parameter-dependent convergence rates, self-normalization, subsampling confidence intervals

Programado

JC5 Series temporales 1
19 de abril de 2012  12:00
Sala Viena

Robust functional classification for time series

A. M. Alonso Fernandez, D. Casado de Lucas, S. Lopez, J. Romo

Simulation tools for statistical model comparison: an application to unobserved component models versus dynamic regression models

B. Dadashova, B. Arenas, J. M. Mira, F. Aparicio Izquierdo

Minimally conditioned likelihood for a nonstationary state space model

S. Sotoca López, M. Jerez Méndez, J. Casals Carro, A. García Hiernaux

Fichero de la comunicación: