Estimating the system order by subspace methods

Issued date: 2007-01

Serie-No.: UC3M Working papers. Statistics and Econometrics07-03

Keywords: System order , State-space models , Subspace methods , Information criteria , Seasonality

Abstract:This paper discusses how to determine the order of a state-space model. To do so, we start byrevising existing approaches and find in them three basic shortcomings: i) some of them have apoor performance in short samples, ii) most of them are not robust anThis paper discusses how to determine the order of a state-space model. To do so, we start byrevising existing approaches and find in them three basic shortcomings: i) some of them have apoor performance in short samples, ii) most of them are not robust and iii) none of them canaccommodate seasonality. We tackle the first two issues by proposing new and refined criteria.The third issue is dealt with by decomposing the system into regular and seasonal sub-systems.The performance of all the procedures considered is analyzed through Monte Carlo simulations.+-

Author: García-Hiernaux, Alfredo; Casals, José; Jerez, Miguel

Source: http://e-archivo.uc3m.es

Teaser

Statistics and Econometrics.
WS 2007-01 Estimating the system order by subspace methods García-Hiernaux, Alfredo http:--hdl.handle.net-10016-932 Descargado de e-Archivo, repositorio institucional de la Universidad Carlos III de Madrid Working Paper 07-03 Statistic and Econometric Series 01 January 2007 Departamento de Estadística Universidad Carlos III de Madrid Calle Madrid, 126 28903 Getafe (Spain) Fax (34-91) 6249849 ESTIMATING THE SYSTEM ORDER BY SUBSPACE METHODS∗ Alfredo García-Hiernaux, 1 José Casals2 and Miguel Jerez 3 Abstract This paper discusses how to determine the order of a state-space model.
To do so, we start by revising existing approaches and find in them three basic shortcomings: i) some of them have a poor performance in short samples, ii) most of them are not robust and iii) none of them can accommodate seasonality.
We tackle the first two issues by proposing new and refined criteria. The third issue is dealt with by decomposing the system into regular and seasonal sub-systems. The performance of all the procedures considered is analyzed through Monte Carlo simulations. Keywords: System order, State-space models, subspace methods, information criteria, seasonality ∗ This research was supported by the Spanish Ministry of Education and Science, under grant SEJ200507388. 1 Department of Statistics, Universidad Carlos III de Madrid, Campus de Getafe, 28903 Madrid (SPAIN). email: aghierna@est-econ.uc3m.es 2 Department of Quantitative Economy, Universidad Complutense de Madrid 3 Department of Quantitative Economy, Universidad Complutense de Madrid 1 Introduction The order of a linear system is the number of dynamic components that must be combined to represent the data dynamics.
Determining this value, also known as McMillan index, is critical in applied data modelling and, accordingl...