home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!portal!lll-winken!unixhub!linac!pacific.mps.ohio-state.edu!zaphod.mps.ohio-state.edu!uakari.primate.wisc.edu!ames!ncar!noao!amethyst!organpipe.uug.arizona.edu!olympus.as.arizona.edu!freeman
- From: freeman@olympus.as.arizona.edu (Jonathan Freeman)
- Newsgroups: sci.math.stat
- Subject: Need Reference for Iteratively Reweighted LS
- Message-ID: <1992Oct15.183239.1619@organpipe.uug.arizona.edu>
- Date: 15 Oct 92 18:32:39 GMT
- Sender: news@organpipe.uug.arizona.edu
- Distribution: usa
- Organization: University of Arizona, Tucson, AZ
- Lines: 13
-
- I have been using iteratively reweighted least squares to fit a non-linear
- model with a number of parameters. The errors are unequal so I use
- weighted least squares. The problem I have is that I am unsure of the
- final scaling factor. Usually this scaling factor involves a square and
- derivative of the influence function used. I have a number of papers (Huber, Street et al) which discuss this, but they are all a bit fuzzy. Can
- anyone give me a VERY straightforward reference preferably in
- a journal. Again I want a good reference for weighted multi-parameter
- IRLS and am unsure of how to obtain the final errors on the parameter
- estimates. Thanks! e-mail or post.
-
- Jonathan Freeman
- Steward Observatory
- University of Arizona
-