On-line estimation of ARMA models using Fisher-scoring

Abdelhamid Ouakasse and Guy Melard

ULB Institutional Repository from ULB -- Universite Libre de Bruxelles

Abstract: Recursive estimation methods for time series models usually make use of recurrences for the vector of parameters, the modelerror and its derivatives with respect to the parameters, plus a recurrence for the Hessian of the model error. An alternativemethod is proposed in the case of an autoregressive-moving average model, where the Hessian is not updated but is replaced,at each time, by the inverse of the Fisher information matrix evaluated at the current parameter. The asymptotic properties,consistency and asymptotic normality, of the new estimator are obtained. Monte Carlo experiments indicate that the estimatesmay converge faster to the true values of the parameters than when the Hessian is updated. The paper is illustrated by anexample on forecasting the speed of wind.

Keywords: ARMA processes; Fisher information matrix; Recursive estimation; Time series (search for similar items in EconPapers)
Date: 2014
New Economics Papers: this item is included in nep-ecm, nep-ets and nep-for
Note: SCOPUS: ar.j
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Published in: Systems research and information science (2014) v.2 n° 1,p.406-432

Downloads: (external link)
https://dipot.ulb.ac.be/dspace/bitstream/2013/1384 ... _2E2014_2E912572.pdf 21642583_2E2014_2E912572 (application/pdf)
https://dipot.ulb.ac.be/dspace/bitstream/2013/13844/1/Ouakasse_Melard_2014.pdf Ouakasse_Melard_2014 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ulb:ulbeco:2013/13844

Ordering information: This working paper can be ordered from
http://hdl.handle.ne ... ulb.ac.be:2013/13844

Access Statistics for this paper

More papers in ULB Institutional Repository from ULB -- Universite Libre de Bruxelles Contact information at EDIRC.
Bibliographic data for series maintained by Benoit Pauwels ().