TY - JOUR

T1 - Integer valued AR processes with explanatory variables

AU - Enciso-Mora, Victor

AU - Neal, Peter John

AU - Subba Rao, Tata

PY - 2009

Y1 - 2009

N2 - Integer valued AR (INAR) processes are perfectly suited for modelling countdata. We consider the inclusion of explanatory variables into the INARmodel to extend the applicability of INAR models. This greatly extends therange of time series data sets to which INAR models can be applied andoffers an alternative to Poisson regression models. An efficient MCMC algorithmis constructed to analyze the model and incorporates both explanatoryvariable and order selection. The applicability of the methodology is demonstratedby considering three different data sets; monthly polio incidences inthe USA 1970-1983, monthly benefit claims from the logging industry to theBritish Columbia Workers’ Compensation Board 1985-1994 and the dailyscore achieved by a schizophrenic patient in a test of perceptual speed.

AB - Integer valued AR (INAR) processes are perfectly suited for modelling countdata. We consider the inclusion of explanatory variables into the INARmodel to extend the applicability of INAR models. This greatly extends therange of time series data sets to which INAR models can be applied andoffers an alternative to Poisson regression models. An efficient MCMC algorithmis constructed to analyze the model and incorporates both explanatoryvariable and order selection. The applicability of the methodology is demonstratedby considering three different data sets; monthly polio incidences inthe USA 1970-1983, monthly benefit claims from the logging industry to theBritish Columbia Workers’ Compensation Board 1985-1994 and the dailyscore achieved by a schizophrenic patient in a test of perceptual speed.

M3 - Journal article

VL - 71

SP - 248

EP - 263

JO - Sankyha B : Applied and Interdisciplinary Statistics

JF - Sankyha B : Applied and Interdisciplinary Statistics

SN - 0976-8394

IS - 2

ER -