EcoSta 2019: Start Registration
View Submission - EcoSta2019
A0299
Title: Opening up the black box: Gaussian process modeling using information from partial differential equation models Authors:  Matthias Tan - City University of Hong Kong (Hong Kong) [presenting]
Abstract: Gaussian process (GP) emulators of computer models are typically constructed based purely on data from a computer experiment using a standard stationary GP prior with product Matern or Gaussian correlation function. This often ignores valuable engineering and mathematical knowledge about the behavior of the computer model. We will focus on the use of known behavior/properties of partial differential equation models solved numerically by computer codes to improve construction of GP emulators for this type of computer models.