Perfect classifiers in partial observability bivariate probit

There have been numerous applications of partial observability bivariate probit models. These models identify determinants of individual discrete outcomes when all that is observed are collective outcomes and individual characteristics. Numerical difficulties with their likelihood functions are commonly reported. This note looks at one aspect: the existence of perfect classifiers.

MoreLess

Year of publication:	2012
Authors:	Poirier, Dale J.
Published in:	Economics Letters. - Elsevier, ISSN 0165-1765. - Vol. 116.2012, 3, p. 361-362
Publisher:	Elsevier
Subject:	Bayesian \| Maximum likelihood \| Prior

More details

Type of publication:	Article
Classification:	C1 - Econometric and Statistical Methods: General ; C2 - Econometric Methods: Single Equation Models ; C11 - Bayesian Analysis
Source:	RePEc - Research Papers in Economics

Persistent link: https://ebvufind01.dmz1.zbw.eu/10010594167