Bayesian Semiparametric Seemingly Unrelated Regression

Abstract

Parametric seemingly unrelated regression (SUR) models are a common tool for multivariate regression analysis when error variables are reasonably correlated. A weakness of parametric models is that they require strong assumptions on the functional form of possibly nonlinear effects of metrical covariates. In this paper, we develop a semiparametric SUR model based on Bayesian P-splines. Inference is fully Bayesian and uses recent Markov chain Monte Carlo techniques.