Important features of multivariate linear regression are emphasised and a selection of prior distributions discussed. Priors used by Brown and Zidek (1978) lead them to a class of 'empirical' Bayes shrinkage estimates. The strength of shrinkage is examined with respect to an election forecasting example where observations obtain one after another.
@article{urn:eudml:doc:40824,
title = {Aspects of multivariate regression.},
journal = {Trabajos de Estad\'\i stica e Investigaci\'on Operativa},
volume = {31},
year = {1980},
pages = {249-265},
language = {en},
url = {http://dml.mathdoc.fr/item/urn:eudml:doc:40824}
}
Brown, Philip J. Aspects of multivariate regression.. Trabajos de Estadística e Investigación Operativa, Tome 31 (1980) pp. 249-265. http://gdmltest.u-ga.fr/item/urn:eudml:doc:40824/