333 articles – 69 Notices  [english version]
HAL : hal-00643089, version 1

Fiche détaillée  Récupérer au format
A parametric study for the first-order signed integer-valued autoregressive process
Christophe Chesneau 1, Maher Kachour 2
(21/11/2011)

In recent years, many attempts have been made to find accurate models for integer-valued times series. The SINAR (for Signed INteger-valued AutoRegressive) process is one of the most interesting. Indeed, the SINAR model allows negative values both for the series and its autocorrelation function. In this paper, we focus on the simplest SINAR(1) model under some parametric assumptions. Explicitly, we obtain the form the probability mass function of the innovation when the marginal distribution of the process is known. Moreover, we give an implicit form of the stationary distribution for a known innovation. Simulation experiments as well as analysis of real data sets are carried out to attest the models performance.
1 :  Laboratoire de Mathématiques Nicolas Oresme (LMNO)
CNRS : UMR6139 – Université de Caen Basse-Normandie
2 :  Laboratoire de Mathématiques Jean Leray (LMJL)
CNRS : UMR6629 – Université de Nantes – École Centrale de Nantes
Laboratoire de Mathématiques Jean Leray, Université de Nantes
Mathématiques/Statistiques

Statistiques/Théorie
Integer-valued time series – INAR models – SINAR models – Rademacher$(p)-\mathbb{N}$ class – Skellam distribution
Liste des fichiers attachés à ce document : 
PDF
P-SINAR.pdf(798.1 KB)