ADDITIVE COVARIANCE KERNELS FOR HIGH-DIMENSIONAL GAUSSIAN PROCESS MODELING

Nicolas Durrande, David Ginsbourger, Olivier Roustant

arXiv (Cornell University) · 2011 · 56 citations · 10 references

DOIFull text

Open access

Concepts

Abstract

Gaussian process models -also called Kriging models- are often used as\nmathematical approximations of expensive experiments. However, the number of\nobservation required for building an emulator becomes unrealistic when using\nclassical covariance kernels when the dimension of input increases. In oder to\nget round the curse of dimensionality, a popular approach is to consider\nsimplified models such as additive models. The ambition of the present work is\nto give an insight into covariance kernels that are well suited for building\nadditive Kriging models and to describe some properties of the resulting\nmodels.\n

References

10