Correct new (test) kernel values, via correcting the augmented kernel matrix. Internal use only.
correctionAugmentedKernelVector(k, object, x)
vector of augmented, corrected kernel values
new kernel value vector
a modelKriging fit
new samples (belonging to kernel values k)
Martin Zaefferer and Thomas Bartz-Beielstein. (2016). Efficient Global Optimization with Indefinite Kernels. Parallel Problem Solving from Nature-PPSN XIV. Accepted, in press. Springer.