CMStatistics 2018: Start Registration
View Submission - CFE
Title: Efficient spatial designs for monitoring environmental phenomenon Authors:  Pierre Druilhet - Universite Clermont Auvergne (France) [presenting]
Sylvain Coly - Universite Clermont Auvergne (France)
Abstract: Gaussian processes are often used as surrogate models for computer experiments or as current prior knowledge of spatio-temporal phenomenon. Space-filling designs aim to provide accurate estimate in the whole space. However, in a monitoring context, it is more relevant to focus on crucial areas. We propose three strategies to build spatial designs, depending on the purpose. The first one aim to focus on areas with high expected values, the second one on areas exceeding a given threshold and the third one on level set detection. We also propose some algorithms to construct efficient designs and we compare our designs with space-filling designs.