Penalized Likelihood Kriging Model for Analysis of Computer Experiments

Kriging is a popular metamodeling technique for analysis of computer experiment. However, the likelihood function near the optimum is flat in some situations, and this may lead to very large random variation in the maximum likelihood estimate. To overcome this difficulty, a penalized likelihood approach is proposed for the kriging model. The proposed method is particularly important in the context of a computationally intensive simulation model where the number of simulation runs must be kept small. We demonstrate the proposed approach for the reduction of piston slap, an unwanted engine noise due to piston secondary motion. Issues related to practical implementation of the proposed approach are discussed.