An overview of kriging for researchers

Rodolphe Le Riche (1, 2, 3, 4) , Nicolas Durrande (5)

1 Mines Saint-Étienne MSE - École des Mines de Saint-Étienne
2 LIMOS - Laboratoire d'Informatique, de Modélisation et d'Optimisation des Systèmes
3 FAYOL-ENSMSE - Institut Henri Fayol
4 FAYOL-ENSMSE - Département Génie mathématique et industriel
5 PROWLER.io, 72 Hills Road, Cambridge

This document corresponds to a 2h class on kriging, also known as conditional Gaussian Processes (GP) or Gaussian Process Regression (GPR). The class was part of the MNMUQ2019 summer school on uncertainty quantification. In the first part of the class, we introduce basics of GPR up to a point where it could be coded by a motivated student. The second part of the class covers research topics such as kernel design and interpolation in a RKHS.

Mots clés: kriging; gaussian process regression; interpolation; RKHS

Domaines: Optimisation et contrôle [math.OC]; Ingénierie, finance et science [cs.CE]; Génie mécanique [physics.class-ph]

Rodolphe Le Riche, Nicolas Durrande. An overview of kriging for researchers. Doctoral. Porquerolles, France. 2019. ⟨cel-02285439v1⟩