Model selection by multiple test procedures
- 1 January 1988
- journal article
- research article
- Published by Taylor & Francis in Statistics
- Vol. 19 (1) , 39-44
- https://doi.org/10.1080/02331888808802068
Abstract
A multiple test procedure for inferring the dimension of a general finite parameter model is proposed which consists of individual tests of each of these parameters. If the critical limits of the individual tests are allowed to depend on the sample size in an appropriate way, the test procedure provides a weakly consistent estimate of the minimal 'correct5 subset of the model parametersKeywords
This publication has 12 references indexed in Scilit:
- The behaviour of the Lagrangian multiplier test in testing the orders of an ARMA-modelMetrika, 1985
- Order Estimation in ARMA-Models by Lagrangian Multiplier TestsThe Annals of Statistics, 1983
- Estimating Regression Models of Finite but Unknown OrderInternational Economic Review, 1981
- The Estimation of the Order of an ARMA ProcessThe Annals of Statistics, 1980
- Selection of RegressorsInternational Economic Review, 1980
- Comments on ‘ On model structure testing in system identification’International Journal of Control, 1978
- Weak and strong consistency of the least squares estimators in regression modelsProbability Theory and Related Fields, 1976
- A new look at the statistical model identificationIEEE Transactions on Automatic Control, 1974
- Some recent advances in time series modelingIEEE Transactions on Automatic Control, 1974
- Fitting autoregressive models for predictionAnnals of the Institute of Statistical Mathematics, 1969