Chapter 4. Probabilistic models and probabilistic programs
This chapter covers
- The definition of a probabilistic model
- How probabilistic models are used to answer queries
- The ingredients of a probabilistic model, including variables, dependencies, functional forms, and numbers
- How a probabilistic program represents the ingredients of a probabilistic model
Part 1 of this book introduced probabilistic programming. You learned that a probabilistic reasoning system uses a probabilistic model to answer queries, given evidence, and that probabilistic programming uses a program to represent the probabilistic model. This part of the book goes deeper into representing probabilistic models. You’ll learn a variety of programming techniques for writing probabilistic programs.
But first you need to develop a basic understanding of probabilistic models and how they’re constructed and used to answer queries. This chapter provides that understanding. You’ve already used some of the ideas intuitively in part 1, but now it’s time to get into the fundamentals.
This chapter elaborates on the themes of chapter 1 in much greater depth, so it’ll be useful if you’ve read that chapter. This chapter also describes how probabilistic programs, and Figaro programs in particular, define probabilistic models, so the basic knowledge of Figaro presented in chapter 2 will be helpful.