[R-pkgs] New package: ROCR (Visualizing classifier performance)

[Tobias Sing]

Dear R users,

we are glad to announce the release of our new R package ROCR, for visualizing 
the performance of scoring classifiers (available on CRAN). We hope that the 
package might be useful for those of you working on classification problems. 
For details, see the package description below, or the ROCR website: 
http://rocr.bioinf.mpi-sb.mpg.de. You can get a short overview by typing 
'demo(ROCR)'. Any kind of feedback (questions, comments, suggestions, bug 
reports) is very welcome.

Best regards,
  the ROCRs (Tobias Sing, Oliver Sander, Niko Beerenwinkel, Thomas Lengauer)

Package description:
-------------------------
ROC graphs, sensitivity/specificity curves, lift charts, and precision/recall 
plots are popular examples of trade-off visualizations for specific pairs of 
performance measures. ROCR is a flexible tool for creating 
cutoff-parametrized 2D performance curves by freely combining two from over 
25 performance measures (new performance measures can be added using a 
standard interface). Curves from different cross-validation or bootstrapping 
runs can be averaged by different methods, and standard deviations, standard 
errors or box plots can be used to visualize the variability across the runs. 
The parametrization can be visualized by printing cutoff values at the 
corresponding curve positions, or by coloring the curve according to cutoff. 
All components of a performance plot can be quickly adjusted using a flexible 
parameter dispatching mechanism. Despite its flexibility, ROCR is easy to 
use, with only three commands and reasonable default values for all optional 
parameters.


__________________________________________________________________________
Tobias Sing                         phone: +49 681 9325 315   
Max-Planck-Institut für Informatik  fax  : +49 681 9325 399   
Stuhlsatzenhausweg 85               email: tobias.sing at mpi-sb.mpg.de  
66123 Saarbrücken, Germany          web  : http://www.tobiassing.net