[R] R-Help

Taddesse Kassahun borninmerawi at gmail.com
Sat Mar 28 15:13:06 CET 2009


Greetings and best wishes to you!

Dear all,

I am currently doing my M.Sc. thesis on the combined applications of
robust and ridge regression. I have a set of 6 continuous variables
for more than 35, 000 cases. In fact, I know how to handle ridge
regression and robust regression separately using statistical
packages. However, my data show me that the two problems happen at
once, i.e., both multicollinearity and outliers do exist in my data
set so that I could not able to get an estimator, which is based on
robust-ridge regression using R.

The parameter that I want to estimate is given by: (X'WX + kI)-1 X'WY,
where k is the baising constant in ridge regression and W is the
daigonal matrix of weights, which is obtained by itteratively weighted
least squares.

Thank you for your time and considering my letter!
-- 
    +  +  +  +  +  +  +
“Everyone has a reason to pray”
    Taddesse Kassahun




More information about the R-help mailing list