Feature selection with Random Forest

Random Forest is a supervised model that implements both decision trees and the bagging method. The idea is that the training dataset is resampled according to a procedure called "bootstrap". Each sample contains a random subset of the original columns and is used to fit a decision tree. The number of models and the number of columns are hyperparameters to be optimized. Finally, the predictions of the trees are mixed together calculating the mean value (for regression) or using soft voting (for classification).