s-news
[Top] [All Lists]

Nested Logit Models

To: s-news@lists.biostat.wustl.edu
Subject: Nested Logit Models
From: brandon.paris@kodak.com
Date: Wed, 21 Feb 2001 16:46:15 -0500

From: Brandon L. Paris

Hi --

Does anyone have experience fitting Nested Logit Models as defined in the
text by Ben-Akiva and Lerman (1985) in S-Plus?  These models are used
extensively in civil engineering applications like transportation and
modeling how available modes of transportation will be utilized.  I
presently use SAS to fit "generic", non-nested multinomial logits.  That
package will soon have the capability to fit nested logits, but it will be
part of a module that we don't purchase and I will probably be unable to
justify my company purchasing it.

Estimating these models is not a trivial task, and I'd like to avoid
programming the estimation myself if someone has had experience with it.
If no one has done this in S-Plus, does anyone have information on other
packages that may work (I understand Limpdep and Stata are two
possibilities . . . any idea about available C++ or Fortran code to do
this?).  Thanks.


Brandon L. Paris
brandon.paris@kodak.com
Statistician, Eastman Kodak Company
(716) 724-3064 (KNet 22-43064)
"In God we trust.  All others must bring data."  W. Edwards Deming



<Prev in Thread] Current Thread [Next in Thread>
  • Nested Logit Models, brandon . paris <=