Publication: Problemas de homogeneidad en modelos de series temporales
Loading...
Identifiers
Publication date
1990
Defense date
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Instituto Nacional de Estadística
Impact
Export
Abstract
Este trabajo se ocupa de problemas de homogeneidad en general, y de cambios en el modelo ARMA en particular, de una serie temporal. Dada una serie temporal, se propone un modelo compuesto de diversos elementos. Cada componente puede reflejar cambios de modelo ARMA o de varianza en la perturbación o de ambas cosas a la vez, o heterogeneidades en general. Se propone un método secuencial para detectar y modelizar heterogeneidades en series temporales univariantes y este método se aplica a dos series temporales.--------------------------------------------------------
This work studies homogeneity problems. and in particular ARMA model changes. in time series. For a given time series a multiple component model is proposed. Each component can encompass changes in ARMA model. changes in noise variance or both, as well as more general non-homogeneous behaviour. A modelling procedure is proposed and illustrated by means of two real time series.
This work studies homogeneity problems. and in particular ARMA model changes. in time series. For a given time series a multiple component model is proposed. Each component can encompass changes in ARMA model. changes in noise variance or both, as well as more general non-homogeneous behaviour. A modelling procedure is proposed and illustrated by means of two real time series.
Description
Keywords
Modelos ARMA, Cambio de varianza, Cambio de modelo, Homogeneidad, ARMA models, Variance change, Model change, Homogeinity
Bibliographic citation
Estadística española, v. 32, n. 124, 1990, pp. 417-438