2. Introduction

In this chapter, we first emphasize that we will address learning from data from a physicist’s perspective, which distinguishes it from most data science text. This leads to our first encounter (with many more to follow) of the Bayesian workflow in the physics context, and then to considering machine learning from the physics perspective. Finally, we articulate the virtues we believe are essential for principled Bayesian inference and, more generally, for effective physics research.