Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: RE: discrim command


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: RE: discrim command
Date   Mon, 10 May 2004 20:17:25 +0100

I guess I'd worry about the effects 
of multivariate normality if I knew I had it. 

Nick 
n.j.cox@durham.ac.uk 

SR Millis
 
> Discriminant function analysis may perform better than
> logistic regression in the presence of multivariate
> normality and homogeneity of covariance matrices.

> --- Nick Cox <n.j.cox@durham.ac.uk> wrote:
> > It would be interesting to know
> > under which circumstances this would
> > happen. Experts?
> >
> > Nick
> > n.j.cox@durham.ac.uk
> >
> > Ricardo Ovaldia
> > >
> > > I originally used -logistic- to generate a
> > > "discriminant" function and that worked well (70%
> > > correct classification). However, I wanted to see
> > if I
> > > could do better with a traditional MUV
> > discriminant
> > > procedure.
> >
> > > --- Nick Cox <n.j.cox@durham.ac.uk> wrote:
> > > > Perhaps you can re-formulate your problem
> > > > as a logit regression problem. Numerous
> > > > papers in the literature over 30 years
> > > > or so have argued that a logit framework
> > > > often works as well or better than discriminant
> > > > analysis.

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index