| Date: | Mon, 9 Jul 2007 11:31:11 -0400 |
|---|
| Reply-To: | Ming Chen <chenming@GMAIL.COM> |
|---|
| Sender: | "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU> |
|---|
| From: | Ming Chen <chenming@GMAIL.COM> |
|---|
| Subject: | Re: Solving statistical problems with SAS |
|---|
|
| In-Reply-To: | <1183986018.718761.261790@k79g2000hse.googlegroups.com> |
|---|
| Content-Type: | text/plain; charset=ISO-8859-1; format=flowed |
|---|
Hi Paige,
It is very interesting to know that Proc PLS can solve logistic
regression problems.
Can you please provide some reference web links for this technique?
Thanks
Ming
On 7/9/07, Paige Miller <paige.miller@kodak.com> wrote:
> On Jul 8, 11:37 am, vincent64 <datashap...@gmail.com> wrote:
> > The problem is some form of PLS logistic regression. Can SAS solve it?
> > How much does it cost (do I need to buy SAS Enterprise miner? Data set
> > is very large, 5 MM rows, 80 initial variables)?
> >
> > See description athttp://datashaping.com/contest14004.shtml
>
> SAS PROC PLS (which is part of SAS/STAT) can solve logistic regression
> problems. The response(s) needs to be properly transformed (using the
> logistic transformation) for it to work.
>
> --
> Paige Miller
> paige\dot\miller \at\ kodak\dot\com
>
|
