AIC or SIC in GeoDa Space

[Kate Peng]

Dear all,

I noticed that the results of spatial lag/error model in Geoda Space do not include AIC or SIC values? 

How do we estimate these values in GeoDa Space?

Any feedback is helpful. Thank you.

Regards,

Kate

[Luc Anselin]

AIC and SC are included for the maximum likelihood estimation only. They are not

included for GMM estimation, since those methods do not assume normality (nor a

likelihood).

Hope this helps.

L.

Luc Anselin, PhD

Stein-Freiler Distinguished Service Professor of Sociology and the College

Director, Center for Spatial Data Science

Senior Fellow, NORC

University of Chicago

http://spatial.uchicago.edu 

--
You received this message because you are subscribed to the Google Groups "Openspace List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openspace-lis...@googlegroups.com.
To post to this group, send email to openspa...@googlegroups.com.
Visit this group at https://groups.google.com/group/openspace-list.
For more options, visit https://groups.google.com/d/optout.

[蒂菁]

Dear Luc,

Thank you for clarification.

I would like to know: if SIC/AIC are not applicable in GMM estimation, which criteria could be used to judge which spatial regression performs better? (In my study, I would like to compare the performance of spatial lag, spatial error and SARAR models given the GMM setting).

Any feedback is appreciated. Thank you.

Regards,

Kate

2016-11-08 0:19 GMT+08:00 Luc Anselin <lans...@gmail.com>:

AIC and SC are included for the maximum likelihood estimation only. They are not

included for GMM estimation, since those methods do not assume normality (nor a

likelihood).

Hope this helps.

L.

Luc Anselin, PhD

Stein-Freiler Distinguished Service Professor of Sociology and the College

Director, Center for Spatial Data Science

Senior Fellow, NORC

University of Chicago

http://spatial.uchicago.edu 

On Nov 7, 2016, at 12:57 AM, Kate Peng <katep...@gmail.com> wrote:

Dear all,

I noticed that the results of spatial lag/error model in Geoda Space do not include AIC or SIC values? 

How do we estimate these values in GeoDa Space?

Any feedback is helpful. Thank you.

Regards,

Kate

--
You received this message because you are subscribed to the Google Groups "Openspace List" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openspace-list+unsubscribe@googlegroups.com.
To post to this group, send email to openspace-list@googlegroups.com.

Visit this group at https://groups.google.com/group/openspace-list.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Openspace List" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openspace-list+unsubscribe@googlegroups.com.
To post to this group, send email to openspace-list@googlegroups.com.

[Luc Anselin]

There isn’t really a good measure of fit since the presence of spatial dependence

runs counter to the equal weights assumption implied in something like an R^2.

In GeoDaSpace, we report a pseudo-R^2 which is simply the squared correlation

between observed and predicted. It’s imperfect, but usually good enough to distinguish

a “bad” model from a better one. Ideally, one would like to do something like a cross-validation,

but with spatially dependent data that is difficult to implement.

Hope this helps.

L.

Luc Anselin, PhD

Stein-Freiler Distinguished Service Professor of Sociology and the College

Director, Center for Spatial Data Science

Senior Fellow, NORC

University of Chicago

http://spatial.uchicago.edu 

To unsubscribe from this group and stop receiving emails from it, send an email to openspace-lis...@googlegroups.com.
To post to this group, send email to openspa...@googlegroups.com.

[蒂菁]

Dear Luc,

Thank you so much for clarification. 

Your reply is deeply appreciated. Thank you.

Regards,

Kate

To post to this group, send email to openspa...@googlegroups.com.
Visit this group at https://groups.google.com/group/openspace-list.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Openspace List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openspace-list+unsubscribe@googlegroups.com.

To post to this group, send email to openspa...@googlegroups.com.
Visit this group at https://groups.google.com/group/openspace-list.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Openspace List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openspace-list+unsubscribe@googlegroups.com.
To post to this group, send email to openspace-list@googlegroups.com.
Visit this group at https://groups.google.com/group/openspace-list.
For more options, visit https://groups.google.com/d/optout.