Hi ALL,
I have done a number of interpolations using kriging and idw with gstat,
and use krige.cv to produce the cross-validation (nfold=5) as well.
How can I evaluate the results of cross-validation for individual
variables based on the me, mspe, rmse, Mean square normalized error, r
(observed vs predicted) and so on?
I mean what kind of criteria I should use in my case. For example, with
512 points, a Sph model was developed and resultantly one variable's
cross-validation produced me=-0.04, rmse=0.63, Mean square normalized
error=1.03, and r=0.36. Is the interpolation accepted and how and why?
Cheers
Yong Li
[[alternative HTML version deleted]]
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
