you do not give many details but if you have a multivariate system you
can do:
a) multivariate estimation/simulations with continuous variables.
b) multivariate indicator estimation/simulations (that can include
continuous variables to)
Then compare the results by cross validation. Notice that if your binary
logistic model have a nice classification table (no large error) and
your multivariate estimation of continuous variables is accurate the: is
possible that the option a) give good results because with indicators
usually you loss a lot of resolution in your data and is too sensible to
data density and distribution.
cross validation error is a robust measure of your estimation...
regard
Adrian Martinez
CFSG student
Centre de Geostatistique
Ecole des Mines de Paris
Tib escribi�:
Hi folks,
I need suggestions on this simple question. I have built a logistic
regression model for predicting presence/absence for some kind of
species in a large area. The model does not incorporate any spatial
effect or longtitude or latitude. From this model I can get confidence
intervals for pridictions at all pixels in that area. My concern is,
can we measure the change of confidence against the distance to
observed points? Generally, the closer new observations to existed
ones (which are used to build the model), the higher confidence we
have for the pridictions because of intrinsic spatial autocorrelation.
Does variogram do this kind of job? Or any other references? Thank you
very much.
------------------------------------------------------------------------
* By using the ai-geostats mailing list you agree to follow its rules
( see http://www.ai-geostats.org/help_ai-geostats.htm )
* To unsubscribe to ai-geostats, send the following in the subject or in the
body (plain text format) of an email message to [EMAIL PROTECTED]
Signoff ai-geostats
* By using the ai-geostats mailing list you agree to follow its rules
( see http://www.ai-geostats.org/help_ai-geostats.htm )
* To unsubscribe to ai-geostats, send the following in the subject or in the
body (plain text format) of an email message to [EMAIL PROTECTED]
Signoff ai-geostats
