Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: RE: RE: ACE algorithm in STATA ?


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: RE: RE: ACE algorithm in STATA ?
Date   Thu, 8 Jun 2006 23:34:50 +0100

You are correct that we quote this paper in the
on-line help and again in the paper in Stata Journal
5(3), 2005. But there is no relevance beyond what
is mentioned, a reference to the guaranteed convergence of
the backfitting algorithm. In particular, -mrunning- 
is not an implementation of ACE as defined by Breiman
and Friedman. It does no transformations, 
and it does no search for transformations. It is 
a smoothing program, rather than a regression program in
the classical sense. It is thus not accurate to say 
that -mrunning- is "based on" Breiman and Friedman's 
paper. (I would have said Yes to your original question
otherwise.) 

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

Lei Xuan
 
> Hi Nick,
> 
> I noticed that -mrunning- was implemented by Patrick Royston 
> and you, based
> on Breiman and Friedman's paper. How is -mrunning- relevant to ACE
> algorithm?
 
Nick Cox
 
> . findit alternating conditional expectations 
> 
> indicates that the answer is no so far as Stata 
> is concerned, and my memory is consistent with 
> that. But a single counter-example would refute 
> it. 
 
Lei Xuan
  
> > I would like to know whether ACE (alternating conditional 
> expectation)
> > algorithm has been implemented in STATA? ACE algorithm was 
> proposed by
> > Breiman and Friedman (1985) in their paper:
> > 
> >  Breiman, L. and Friedman, J. H. (1985). " Estimating optimal
> > transformations for multiple regression and correlation". J. 
> > Amer. Statist.
> > Assoc. 80, 580.

*
*   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