[R-pkgs] caret package version 4.63

[KuhnA03]

Version 4.63 of the caret package is now on CRAN.

caret can be used to tune the parameters of predictive models using
resampling, estimate variable importance and visualize the results.
There are also various modeling and "helper" functions that can be
useful for training models.

caret has wrappers to over 99 different models for classification
and regression. See the package vignettes or:

  http://user2010.org/slides/Kuhn.pdf
  http://www.jstatsoft.org/v28/i05

for more details. 


Since the last posting to the list:

 - wrappers for a number of new models were added, notably gam
   models (from both the gam and mgcv packages) and logic trees

 - when resampling with train(), class probabilities can now be
   used to calculate performance (such as the AUC of an ROC curve).
   A basic summary function, twoClassSummary(), can be used to
   calculate sensitivity, specificity and the ROC AUC.

 - repeated k-fold CV and the bootstrap 632 technique are available
   in train()

 - pre-processing can not be used within each resampling iteration
   within train().

 - a function for independent component regression (icr) was added

 - the class for aggregating and visualization resampling results
   (resamples) has been enhanced with more visualization methods.
   The class can also work with caret's feature selection routines
   (rfe() and sbf())

 - the print method for train() has been improved

 - functions can be now be passed to the tuneGrid argument in
   train()

 - an existing function that catalogs the existing models available
   within train(), called modelLookup(), is now available to the
   users

 - when parallel processing, more computations are being executed
   in the worker processes than previously (e.g. performance calcs)

The NEWS file has the blow-by-blow list of changes.


The package homepage is

  https://r-forge.r-project.org/projects/caret/

Send questions, collaborations, comments etc to max.kuhn at pfizer.com.

Max