Structure identification of parameter-bounding models by use of noise-structure bounds

1 August 1989

journal article
research article
Published by Taylor & Francis in International Journal of Control

Vol. 50 (2) , 639-649
https://doi.org/10.1080/00207178908953386

Abstract

The selection of a model for a given set of records is discussed for the case when the output error is required to be always within some specified bounds. A new selection criterion is proposed: the model should be the simplest giving sufficiently unstructured output errors. The lack of structure is specified by requiring the candidate model to be capable of giving sufficiently small sample autocorrelations of the model-output errors over a range of lags. Computational algorithms to apply the criterion are presented, and illustrated by tests on AR processes.

Keywords

This publication has 7 references indexed in Scilit:

Recursive Estimation of Autoregressions
Journal of the Royal Statistical Society Series B: Statistical Methodology, 1989
Identification and application of bounded-parameter models
Automatica, 1987
Multivariate linear time series models
Advances in Applied Probability, 1984
Order Estimation in ARMA-Models by Lagrangian Multiplier Tests
The Annals of Statistics, 1983
On the value of information in system identification—Bounded noise case
Automatica, 1982
Selection of the order of an autoregressive model by Akaike's information criterion
Biometrika, 1976
A new look at the statistical model identification
IEEE Transactions on Automatic Control, 1974