One possibility that I can think of is Friedman's MART. There's no
implementation in Splus AFAIK. However, Greg Ridgeway has an R package
called "gbm" that has similar functionality. If you can use R, you may want
to give that a try.
HTH,
Andy
> -----Original Message-----
> From: Nancy Friday [mailto:Nancy.Friday@noaa.gov]
> Sent: Monday, September 29, 2003 9:18 PM
> To: S-News Mailing List
> Subject: [S] GAM with a multinomial response
>
>
> Hi,
>
> I'm looking for a way to model multinomial data (both nominal and
> ordinal) as a function of multinomial and continuous
> explanatory variables in a generalized additive framework.
> Is there an easy way to do this in splus?
>
> Thanks,
> Nancy
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> * Nancy Friday, Ph.D.
> * Research Fishery Biologist
> * Cetacean Assessment and Ecology Program
> * National Marine Mammal Lab, AFSC/NMFS
> * 7600 Sand Point Way NE
> * Seattle, WA 98115-6349 USA
> * +1 206-526-6266
> * +1 206-526-6615 (fax)
> * Nancy.Friday@noaa.gov
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
> --------------------------------------------------------------------
> This message was distributed by
> s-news@lists.biostat.wustl.edu. To unsubscribe send e-mail
> to s-news-request@lists.biostat.wustl.edu with the BODY of
> the message: unsubscribe s-news
>
|