Multinomial Logistic Regression

YI-MEI.ZIA@firstunion.com
Tue, 13 Jan 1998 19:42:37 -0500


Date: Tuesday, 13 January 1998 9:36am ET
To: Softswitch.OAS
Cc: *
From: YI-MEI.ZIA@TAOFUN1
Subject: Multinomial Logistic Regression


I did have an type error on my previous email. I went to the right signt
had got something like this:
Multivariate Statistics


Classification and Clustering

classif
dysect
mclust (in S-PLUS)
multiv

But did not find anyway to go further. Please help]

My previous email wrote:

After I read the following attached mail, I have tried to search fo
r more info. about the Multinomial Logistic Regression on the inter
net but it is hard to find it without a address. The site that I we
nt into is http://lib.stat/cmu.edu/S/topical.index. Please advise m
e the correct site and library. Thank you ]]

Date: Wednesday, 24 December 1997 6:14pm
To: s-news
To: KKALYANAM
From: INTERNET.SNEWSREQ@SSWITCH.SNADS
Subject: Re: Multinomial Logit Estimation

> From: Kirthi Kalyanam <KKALYANAM@mailer.scu.edu>
> Does anybody know of/or has written a function for multinomial logit
> estimation in splus.

(1) multinom (by Brian Ripley), available from Statlib, for the basic
multinomial regression model.
(2) Design (by Frank Harrell), also from Statlib, for ordered multinomial
regression models.
(3) plr, available from me (not yet posted to Statlib), for multinomial
regression with structured (e.g. cross-classified) outcomes.