[ai-geostats] validation with gslib
Working with GSlib i wondered if anybody ever adjusted the simulation
procedures (sgsim, iksim, dtsim...) so you can actually perform an
independent validation? So not crossvalidation off course, but in my
opinion a jackkniving should be possible.
If you could first simulate the complete grid and afterwards simulate
the external points it seems to me like a perfect valid jackkniving
procedure? Suggestions or remarks? Anyone actually adjusted the source
code to perform such a validation?
- Dear list
I have a question regarding sampling terminology, and especially
Second-phase sampling occurs when there is a need/desire to augment an
initial set of measurements to improve results/gain confidence. We denote
the initial sampling set M, and the second-phase sampling set N.
The set M has m elements, the set N has n elements. Those n new samples can
be added two ways:
(1) One-by-one, n-times and this is referred to as SEQUENTIAL ADDITION
(2) All in once.
My question is whether there is a correct terminology for the second case.
Thank you much.
Best, Eric Delmelle