[R] Question regarding GBM package

Max Kuhn mxkuhn at gmail.com
Fri May 21 15:00:19 CEST 2010

The caret package can do a lot of that for you:



On Fri, May 21, 2010 at 8:05 AM, Roel Meeuws <r.j.meeuws at tudelft.nl> wrote:
> Dear R expert
> I have  come across the GBM package for R and it seemed appropriate for my
> research. I am trying to predict the number of FPGA resources required by a
> Software Function if it were mapped onto hardware. As input I use software
> metrics (a lot of them). I already use several regression techniques, and
> the graphs I produce with GBM look promising.
> Now my question... I see that the output of the GBM package gives (when
> using cross-validation) also an array called cv.error. How might I obtain
> the Cross-Validated Rooted Mean Square Error  from that data? Or is there
> another approach to that?
> Also I would like to have a plot of the cross-validated predictions versus
> the original data, I could do this by manually performing Leave-One-Out and
> getting the predictions for the plot, but as GBM incorporates
> Cross-Validation I was wondering if there is an easier approach.
> I hope someone can point me in the right direction. Many thanks for any help
> anyone might be able to give.
> kind regards,
> Roel Meeuws
> --------------------------------------------
> Roel Meeuws
> PhD. Student
> Delft University of Technology
> Faculty of Electrical Engineering Mathematics and Computer Science
> Computer Engineering Laboratory
> Mekelweg 4, 2628 CD Delft, The Netherlands
> --------------------------------------------
> Email:r.j.meeuws at tudelft.nl <Email%3Ar.j.meeuws at tudelft.nl>
> Office: HB 16.290
> Office phone: +31 (0)15 27 82 165
> Mob. phone: +31 (0)6 10 82 44 01
> --------------------------------------------
>        [[alternative HTML version deleted]]
> ______________________________________________
> R-help at 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.



More information about the R-help mailing list