Dual kriging

Polatory can perform kriging prediction using RBF interpolation, known as dual kriging. Dual kriging solves the same problem as kriging but in a far more efficient way, and both methods should produce the same results. Here is the correspondence between the concepts of the both worlds:

Kriging	RBF interpolation
Prediction	Interpolation (fitting + evaluation)
Covariance function	Positive definite basic function
Generalized covariance function of order k	Conditionally positive definite basic function of order k + 1
Nugget effect model	Regularization/smoothing
Simple kriging	Interpolant with no polynomial
Ordinary kriging	Interpolant with a polynomial of degree 0
Universal kriging with polynomial trend	Interpolant with a polynomial of degree ≥ 1
Weights	(Not computed)
Kriging variance	(Not computed)

For the list of available (generalized) covariance functions, see List of basic functions.

References

1. J. Chilès and P. Delfiner, Geostatistics: Modeling Spatial Uncertainty, Wiley Series in Probability and Statistics (2012). https://doi.org/10.1002/9781118136188