Optimal Model Average Prediction in Orthogonal Kriging Models

The main objective of this paper is to consider model averaging methods for kriging models. This paper proposes a Mallows model averaging procedure for the orthogonal kriging model and demonstrate the asymptotic optimality of the model averaging estimators in terms of mean square error. Simulation studies are conducted to evaluate the performance of the proposed method and compare it with the competitors to demonstrate its superiority. The authors also analyse a real dataset for an illustration.