Topics in Identification, Limited Dependent Variables, Partial Observability, Experimentation, and Flexible Modeling by Jeliazkov Ivan;Tobias Justin;

Author:Jeliazkov, Ivan;Tobias, Justin;
Language: eng
Format: epub
Publisher: Emerald Publishing Limited
Published: 2019-07-18T00:00:00+00:00

The same procedure can be deployed in a maximum-likelihood framework either to economize on the need for MCMC runs or simply because the investigator prefers a frequentist approach. In this section we offer a frequentist version of the BMA in which we mix the maximum-likelihood distributions of the two models, using the Schwarz criterion to compute mixture probabilities. In a sufficiently large sample, Bayesian and frequentist estimators coincide. More interestingly, we find that for the size samples often used in ARMA estimates and with relatively diffuse priors, the Bayesian and frequent results of our procedure are quite similar.

The need for explicit priors and MCMC simulations to estimate the marginal likelihood can be eliminated by use of the Schwarz criterion

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.