Hi I have some data (v. large amount) with a (0,1) response where I want to minimise the errors in the betas rather than SS or deviance. So can anyone point me to a ridge regression function or equivalent for such a logistic regression case? John -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
John Logsdon wrote:> Hi > > I have some data (v. large amount) with a (0,1) response where I want to > minimise the errors in the betas rather than SS or deviance. > > So can anyone point me to a ridge regression function or equivalent for > such a logistic regression case?John, I think you can do that with the nnet code of Brian Ripley. Use the softmax setting and the weight decay regularizer, which is the same as using a ridge penalty term, isn't it? Adrian -- Adrian Trapletti, Vienna University of Economics and Business Ad- ministration, Operations Research, Augasse 2-6, 1090 Vienna, Austria Phone: ++43-(0)1-31336-4561 Email: adrian.trapletti at wu-wien.ac.at Fax: ++43-(0)1-31336-708 WWW: http://quor.wu-wien.ac.at/adrian.html -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
John Logsdon <j.logsdon at lancaster.ac.uk> writes:> > I have some data (v. large amount) with a (0,1) response where I want to > minimise the errors in the betas rather than SS or deviance. > > So can anyone point me to a ridge regression function or equivalent for > such a logistic regression case?John You could start by looking at V&R Examples http://www.stats.ox.ac.uk/pub/MASS3/VR3ans.zip There is a small ridge regression problem there. You will need to adapt it to the glm case by hacking at the standard glm.fit function. Good luck Ross Darnell -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._