Hi Levi, Thank you very much for your reply and concern.
There was a package till 2011...or even 2012...entitled "Generalized Path Seeking" Regression in "R" which was using a combination of all of these methods...as a way to regularize regression. However, I presently can't find it. Thus, I was wondering if someone in the R-community could inform us if there is an upgraded version of it with a different name or another package which does the same thing. Thank you again very much and my apologies for populating the mail messages of the community. Cheers Christos On Wed, Apr 17, 2013 at 9:50 AM, Levi Waldron <lwaldron.resea...@gmail.com>wrote: > Perhaps I am wrong, but I think there are only a few packages supporting > Elastic Net, and none of them also perform Best Subsets. > > > On Wed, Apr 17, 2013 at 8:46 AM, Christos Giannoulis > <cgiann...@gmail.com>wrote: > >> Merhaba, Hello to you too Mehmet (Yasu ki sena) >> >> Thank you for your email and especially for sharing this package. I >> appreciate it. >> >> However, my feeling is that this package does not have the third component >> of Best Subsets (pls correct me if I am wrong). It uses only a combination >> of Ridge and Lasso. >> >> If you happen to know any other packages that uses all of them I would >> greatly welcome and appreciate if you were so kind and share it. I tried >> to >> search the cran lists but I am not sure I can found something like that. >> That's why I was asking the R-community >> >> Thank you again for your prompt response! >> >> Cheers >> >> Christos >> >> On Wed, Apr 17, 2013 at 8:16 AM, Suzen, Mehmet <msu...@gmail.com> wrote: >> >> > Yasu, >> > >> > Try Elastic nets: >> > http://cran.r-project.org/web/packages/pensim/index.html >> > >> > There some other packages supporting elastic nets: Just search the CRAN >> > >> > Cheers, >> > Mehmet >> > >> > >> > On 17 April 2013 13:19, Christos Giannoulis <cgiann...@gmail.com> >> wrote: >> > > Hi all, >> > > >> > > I would greatly appreciate if someone was so kind and share with us a >> > > package or method that uses a regularized regression approach that >> > balances >> > > a regression model performance and model complexity. >> > > >> > > That said I would be most grateful is there is an R-package that >> combines >> > > Ridge (sum of squares coefficients), Lasso: Sum of absolute >> coefficients >> > > and Best Subsets: Number of coefficients as methods of regularized >> > > regression. >> > > >> > > Sincerely, >> > > >> > > Christos Giannoulis >> > > >> > > [[alternative HTML version deleted]] >> > > >> > > ______________________________________________ >> > > R-help@r-project.org mailing list >> > > https://stat.ethz.ch/mailman/listinfo/r-help >> > > PLEASE do read the posting guide >> > http://www.R-project.org/posting-guide.html >> > > and provide commented, minimal, self-contained, reproducible code. >> > >> >> [[alternative HTML version deleted]] >> >> ______________________________________________ >> R-help@r-project.org mailing list >> https://stat.ethz.ch/mailman/listinfo/r-help >> PLEASE do read the posting guide >> http://www.R-project.org/posting-guide.html >> and provide commented, minimal, self-contained, reproducible code. >> > > [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
