POLAR is capable of two distinct learning objectives: regret minimization and active learning. The details for each of these learning objective are provided using the following links.