Introduction
Machine learning never deals with certainty: data are noisy, measurements are incomplete, and predictions can be wrong. Probability is the language for reasoning about uncertainty. This topic sets up its three building blocks — what can happen, which groups of outcomes we care about, and how likely each is.