Descripción
|
|
---|---|
This paper has been developed with the purpose of applying simulation tools to model comparison. Two dynamic statistical models related to road accidents and involving vans have been chosen: Demand for road use, accidents and their gravity (DRAG), developed by Gaudry (1984) and Gaudry and Lasarre (2000) and Unobserved Components Models (UCM) with intervention, proposed by Harvey and Durbin (1986) and further developed by Hermans et al. (2005) and Bijleveld et al. (2008, 2010). The main difference between the two models lies on the fact that, unlike DRAG, the UCM includes specific terms for both trend and seasonal components, which are assumed unobserved, with their corresponding state equations. The central idea is to assume that one of the models is the true one, and study the effect of estimating the other one on parameter interpretation, goodness of fit and prediction accuracy. | |
Internacional
|
No |
Nombre congreso
|
XXXIII Congreso Nacional de Estadística e Investigación Operativa VII Jornadas de Estadística Pública |
Tipo de participación
|
960 |
Lugar del congreso
|
Madrid |
Revisores
|
No |
ISBN o ISSN
|
|
DOI
|
|
Fecha inicio congreso
|
17/04/2012 |
Fecha fin congreso
|
20/04/2012 |
Desde la página
|
86 |
Hasta la página
|
86 |
Título de las actas
|