s-news
[Top] [All Lists]

what type of robust regression technique is behind the SPLUS function "l

To: <s-news@lists.biostat.wustl.edu>
Subject: what type of robust regression technique is behind the SPLUS function "leaps"?
From: "Ana Sebastian" <ana.sebastian@jrc.it>
Date: Thu, 26 Apr 2001 15:05:58 +0200
Dear all,
I would need to find out what type of robust regression technique is behind
the SPLUS function "leaps". In the on-line help one can read: "The leaps
function provides a way to select a few promising regressions (sets of
explanatory variables) for further study.  By using robustness weights found
by a robust regression on the full set of variables, this search can find
suitable regressions even when outliers are present".
Well,  I have been looking in several splus manuals and realised that there
are several robust regression techniques... Could anybody kindly give me
some further information on what exactly is behind this function?
Thank you very much in advance. Just in case it matters I'm using SPLUS 4.5.
Ana.

Ana Sebastian
 EC-DG Joint Research Centre
 Space Applications Institute
 EGEO-Natural Hazards (TP 263)
 I-21020, Ispra (Va), Italy .

 Phone: 39 0332 786376
 Fax:      39 0332 785500
 email: ana.sebastian@jrc.it



<Prev in Thread] Current Thread [Next in Thread>
  • what type of robust regression technique is behind the SPLUS function "leaps"?, Ana Sebastian <=