“…The solution, which minimises the mean squared prediction error, assumes the variable
+Ψ( x ) consists of a regression, in particular a linear combination of m function f ( x ) with linear combination parameters β , and the spatially correlated regression error
, having
the correlation matrix dependent on the distance h and a set of parameters θ . The correction term depends on the residuals, with F is the matrix with entries
and Y n the training set with n data, weighted by the correlation, let
39,45,46 . When training a GP model, it is crucial to determine the regression model F and the correlation parameters.…”