# R package for feature set algorithm selection

I want to train a binary classification NN and part of this will require data pre-processing. However, I have a choice of which pre-processing algorithm to use. Of course I'd like to choose that one which is maximally informative for the purposes of training the NN. I feel sure that there might be some R package that I could us for this selection purpose, so could anyone point me in the right direction?

-
You should cross-validate the entire algorithm, feature selection included. The train function in the caret package offers many pre-processing options, the effects of which are included in the cross validation loop.