Random Forests Lab

Goal

In this lab, we’ll again try to predict players’ skill level, that is, the SkillLevel column, which takes integer values between 0 and 3, with 0 being novice and 3 being expert. We’ll use random forests in this lab. There are many different implementations of random forest (RF) that are wrapped and provided in caret (check out this link for available models incaret). We are going to use the implementation by ranger, since it is reportedly faster than others, and it provides more freedom to tune the parameters.

Brief refresher

Random forest (RF) is an ensemble method—that is, a collection of models that builds upon decision trees—thus the forest in its name. As such, it’s able to capture nonlinearity in the data, is more robust to noise, and is less prone to overfitting.

Preparation