Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: random forest algorithm in Stata?


From   Joseph Luchman <[email protected]>
To   statalist <[email protected]>
Subject   Re: st: random forest algorithm in Stata?
Date   Tue, 7 Jan 2014 10:48:54 -0800

Hi Jeph and Austin,

  I am planning on developing an implementation of a random forest
algorithm that uses the CHAID (CHi-square Automated Interaction
Detection) algorithm (which I recently posted to SSC; type findit
chaid) as the base learners.  In this sense it's kind of like Hothorn,
Hornik, Strobl, and Zeilis' -cforest- algorithm in the R package
-party- that uses conditional inference trees as the base learners
(instead of, I believe, the C4.5 algorithm).

 That said, I'm mainly interested in moving forward with CHAID and
have no plans for implementing the original random forest algorithm
described by Breiman (2001).  I look forward to comparing how the
algorithms perform.

- joe

reference

Leo Breiman (2001). Random Forests. Machine Learning, 45(1), 5–32.

Joseph Nicholas Luchman, M.A.
----
Behavioral Statistics Lead | Fors Marsh Group
Email: [email protected]
forsmarshgroup.com
----
Doctoral Candidate
Industrial Organizational Psychology
George Mason University
https://www.researchgate.net/profile/Joseph_Luchman/

In that case, let me know when you are done, in case I haven't got
started yet...

cheers,
Jeph

On 1/6/2014 10:54 AM, Austin Nichols wrote:

Jeph Herrin <[email protected]> :

Funny, I just asked someone from StataCorp yesterday at the AEA
meetings if they knew of anyone working on an implementation, as I had
one underway and was thinking about submitting it for the Stata Conf
in Boston at the end of July:
http://www.stata.com/meeting/boston14/
(and didn't want to duplicate anyone else's effort).

So as a short answer: I think not, but there soon will be, one way or another.


On Mon, Jan 6, 2014 at 9:38 AM, Jeph Herrin <[email protected]> wrote:

All,

Has anyone implemented any version of the random forest classification
algorithm in Stata?

http://en.wikipedia.org/wiki/Random_forest

I've been using the R packages -randomForest- and -randomSurvivalForest- and
am thinking of writing a Stata implementation. However, would rather not
reinvent if there is something out there already.

I searched SSC, the Statalist archive, and the World Wide Web (c), to no
avail, but still thought I'd ask.

thanks,
Jeph

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index