topic Re: Which model is better in GA? in ArcGIS GeoStatistical Analyst Questions

Which model is better in GA?

Chia_HsunLee — Sun, 26 May 2013 05:31:00 GMT

im using GA to deal with raining data, and so confused on the rule to choose the better model

Rule1
in the book, it said you need choose a model to FIT your semivariogram points

Rule2
in the HELP CENTER, it said smaller Root-Mean-Square , nearest between Root-Mean-Square and Average Standard Error , and
make Root-Mean-Square standardized equal to 1

BUT i got 2 different model results under these 2 rule

Result1 , the model seems very close to the semivariogram trend , but its Statistical property is not obey Rule2

[ATTACH=CONFIG]24678[/ATTACH]

Result2, its Statistical property is exactly accurate on Rule2 , but the model line seems not fit the trend of semivariogram points
[ATTACH=CONFIG]24679[/ATTACH]

so , which one is the "better" model i can use?
or how can i modified to satisified these 2 rules

Re: Which model is better in GA?

EricKrause — Tue, 28 May 2013 14:59:33 GMT

This is a tough call. Both rules are important, but I tend to give more preference to crossvalidation. However, since you have one model with a perfect visual fit and another model with perfect crossvalidation, I think you can find a third model that has both. Try using Spherical and K-Bessel semivariograms.

Re: Which model is better in GA?

Anonymous User — Wed, 29 May 2013 10:21:41 GMT

Original User: van6499

haha very thx

now i found the 3rd model , and got a beautiful result~