Description
RF-ACE, a derivative of Random Forest (RF) and ACE, is a multivariate feature selection algorithm for mixed type categorical and numerical data. RF-ACE assigns p-values for uncovered associations and makes predictions for new observations with Gradient Boosting Trees.