Best Linear Unbiased Prediction: Applications

Abstract

Best linear unbiased prediction (BLUP) can be thought of as best linear prediction (BLP) with one difference: the vector of fixed effects (α) associated with the vector of observations (y) is assumed known in BLP, but estimated in BLUP. Many of the steps required to accomplish BLUP and characteristics of how BLUP handles data are similar to those discussed for BLP in Chapters 4- 8; they are therefore not discussed in detail in this chapter. The purpose of this chapter is to apply BLUP for two simple cases, and to illustrate some interesting aspects of prediction that are unique to BLUP, and therefore not covered in earlier chapters. Finally, we present some examples where use of BLUP might accrue significant advantages over BLP.KeywordsFixed EffectGeneralize InverseGeneralize Little SquareMother TreeProgeny TestThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.