Bayesian comparison of forecasting models to expected progenies difference in Nelore cattle genetic breeding

Abstract

The objective of this work was to accomplish a bayesian analysis of an autoregressive, AR(p), panel data model from Nelore sires' expected progenie difference (EPD) observed during 2000–2006. The AR(2) model was used due to the results of partial autocorrelation function analysis. The prior comparisons were performed through Bayes Factor and Pseudo-Bayes Factor, and the results showed the independent t-Student multivariate – inverse Gamma superiority in relation to the hierarchical multivariate Normal – inverse Gamma and Jeffreys prior. Results indicate the importance of sires grouping by accuracy values, and also show forecast efficiency around 80%.