[S] robust regression

John P McKinlay (john.mckinlay1@jcu.edu.au)
Sat, 31 Jan 1998 19:01:21 +1000


My v3.3 for windows help file for rreg states the following:

"Two ways to currently get high-breakdown, high-efficiency regression are
to use rreg
with weights that are based on the hat diagonal, or to use least squares
with weights
based on the size of the residuals from the least trimmed squares regression."

Can anyone point me toward published material which shows this?

Thanks; please respond to me, I'll summarise for the list.

john mckinlay

-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news