Best linear unbiased prediction

BLUP ( Best Linear Unbiased English abbreviation of Prediction on German best linear unbiased prediction) is a statistical method that can be predicted in multilevel models with random variables. The best linear unbiased estimate non-random variables is, however, called the Best Linear Unbiased Estimator (BLUE). Both variables were introduced in the 1950s by Charles Roy Henderson.

The calculated values ​​with BLUP minimize the estimation error on average ( minimum variance). Calculations are based on mixed model equations in which deterministic and random effects are modeled jointly. BLUP is considered the current best estimation model in the genetic evaluation of animals.