Bayesian Assessment of Dimensionality in Multivariate Reduced Rank Regression
Jukka Corander and Mattias Villani
Abstract
We
consider Bayesian inference about the dimensionality in the multivariate
reduced rank regression framework, which encompasses several models such
as MANOVA, simultaneous equations, factor analysis and cointegration models
for multiple time series. The fractional Bayes approach is used to derive
an approximation to the posterior distribution of the dimensionality.
To investigate the finite sample properties of our solution, we have applied
it to a wide variety of real and simulated data sets. The proposed approach
compares favorably to other established estimators of the dimensionality.