Suman, Im sorry not to have a straight answer. I have never seen PLR used specifically for development of prediction models. However, I could say that there is many many examples around that show no consistent advantage of more modern and complex methods over logistic regression, even if complex non linear relations among the predictors and the predictor towards the outcome are suspected. Thus, you may want to consider other stuff such as software availability and your own experience with a particular method. I suggest these readings:
1. D J Sargent, Comparison of artificial neural networks with other statistical approaches: results from medical data sets, Cancer 91, no. 8 (April 15, 2001): 1636-1642. 1. Peter C. Austin, A comparison of regression trees, logistic regression, generalized additive models, and multivariate adaptive regression splines for predicting AMI mortality, Statistics in Medicine26, no. 15 (7, 2007): 2937-2957. These readings will give the impression that the theories behind these methods bring only theoretical advantages, but do not increase accuracy of prediction when compared to logistic regression. Perhaps, if spend some time to look more around, you may be able to find evidence against this conclusion. Kind regards and good luck with your PhD. Abraço forte e que a força esteja com você, Dr. Pedro Emmanuel A. A. do Brasil Instituto de Pesquisa Clínica Evandro Chagas Fundação Oswaldo Cruz Rio de Janeiro - Brasil Av. Brasil 4365 Tel 55 21 3865-9648 email: pedro.bra...@ipec.fiocruz.br email: emmanuel.bra...@gmail.com ---Apoio aos softwares livres www.zotero.org - gerenciamento de referências bibliográficas. www.broffice.org ou www.openoffice.org - textos, planilhas ou apresentações. www.epidata.dk - entrada de dados. www.r-project.org - análise de dados. www.ubuntu.com - sistema operacional 2010/10/27 <r-sig-epi-requ...@stat.math.ethz.ch> > Send R-sig-Epi mailing list submissions to > r-sig-epi@stat.math.ethz.ch > > To subscribe or unsubscribe via the World Wide Web, visit > https://stat.ethz.ch/mailman/listinfo/r-sig-epi > or, via email, send a message with subject or body 'help' to > r-sig-epi-requ...@stat.math.ethz.ch > > You can reach the person managing the list at > r-sig-epi-ow...@stat.math.ethz.ch > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of R-sig-Epi digest..." > > > Today's Topics: > > 1. Kindly help me of applicability of PLSR method for binary > outcome variable (Suman Kundu) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 27 Oct 2010 01:48:52 -0700 (PDT) > From: Suman Kundu <suman_m...@yahoo.com> > To: r-sig-epi@stat.math.ethz.ch > Subject: [R-sig-Epi] Kindly help me of applicability of PLSR method > for binary outcome variable > Message-ID: <874660.52882...@web112515.mail.gq1.yahoo.com> > Content-Type: text/plain > > Dear friends, > > I am a PhD student at Erasmus Medical Centre, Rotterdam, The Netherlands. I > found the R package "pls", which is very handy and useful. This package > descibes how to use Partial Least Squares Regression (PLSR) and Principal > Component Regression (PCR). > Would you mind to kindly answer the following? > Logistic regression (LR) is a commonly used method in medical science to > make a prediction model for binary outcome variable. This is fine, and I > would like to apply (if applicable) PLSR and PCR to construct a prediction > model for binary outcome variable, and will check whether PLSR or PCR method > helps to make better model compare to LR method, in the sence of > discrimination and goodness-of-fit (calibration). However, I am not sure > whether the R package "pls" allows to construct such a model. > I will be pleased if you kindly let me know whether PLSR ( or PCR) method > is applicable for binary outcome variable. > > Thank you very much for your suggestions. > > Kind regards, > Suman Kundu > > > > [[alternative HTML version deleted]] > > > > ------------------------------ > > _______________________________________________ > R-sig-Epi mailing list > R-sig-Epi@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/r-sig-epi > > > End of R-sig-Epi Digest, Vol 46, Issue 3 > **************************************** > [[alternative HTML version deleted]]
_______________________________________________ R-sig-Epi@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-epi
