Publications scientifiques
Serge Tayim Kelik
* (Auteur correspondant)
Siméon Fotso
Eugène Kouassi
Communications in Statistics - Simulation and Computation
In this paper, the derivation of the Best Linear Unbiased Predictor (BLUP) in the context of the two-way error components model is obtained under fairly general conditions. Then the prediction functions, parameter estimations as well as the gains in efficiency are subsequently derived. The results obtained are extended taking into account autoregressive processes of order 1 (AR(1)) in the error terms. Some Monte Carlo experiments conducted indicate that the model performs reasonably well for some suitable values of N and T.
