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

Re: st: can stata perform ridge regression


From   "Scott Merryman" <[email protected]>
To   <[email protected]>
Subject   Re: st: can stata perform ridge regression
Date   Sun, 14 Mar 2004 17:39:30 -0600

----- Original Message ----- 
From: "Richard Williams" <[email protected]>
To: <[email protected]>; <[email protected]>
Sent: Saturday, March 13, 2004 9:06 AM
Subject: Re: st: can stata perform ridge regression


> At 10:14 PM 2/13/2004 +0800, rghuang wrote:
> >when multicollinearity,it is not suitable to use ols regression.but can
> >use ridge regression,can stata do it?if it can not,how stata deal with
> >multicollinearity?
>
> First off, it is still suitable to use OLS.  Estimates remain
> unbiased.  There is almost always some degree of multicollinearity, so how
> problematic multicollineairity is depends on how extreme it is.
>
> -findit ridge regression- yields a few hits; I can't vouch for how good
> they work.  Greene notes in his Econometric Analysis, 4th edition, p. 258
> that ridge regression produces biased estimators and hence has seen little
> practical use.
>
> For some general discussions on diagnosing and dealing with
> multicollinearity, see
>
> http://www.nd.edu/~rwilliam/xsoc593/lectures/l03.pdf
>
> http://www.ats.ucla.edu/stat/stata/modules/reg/multico.htm
>
> There have also been some recent discussions of multicollinearity on
> Statalist; search the archives.
>

In addition, there was a discussion of ridge regression last November.  The ado
files to preform ridge regession have been updated for version 7+ (but these
will not showup using -findit-).  See the message from May Boggess at
http://www.stata.com/statalist/archive/2003-11/msg00523.html for more details.

Scott

*
*   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 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index