[alsocasey]

The R portion needs some serious filling - for starters: you include the Julia wrapper to glmnet, which was originally implemented in R.

glmnet - lasso/ridge/elastic net glm models.

e1071 - SVM classifiers.

randomForest - random forest classifiers.

mixOmics - a good collection of component-based approaches (PCA, ICA, PLS, etc. includes sparse variants of all of the above is feature selection is required).

caret - similar to Java's Weka.

[mbq]

Or simply a link to the ML CRAN Task View: http://cran.r-project.org/web/views/MachineLearning.html