Bayesian network

A Probabilistic Graphical Model (PGM) that specifically uses a Directed acyclic graph (DAG) to show conditional (causal) dependencies and represent the joint probability function.

Formally is a pair

$$(\mathcal{G},p)$$

where $\mathcal{G}$ is a DAG and $p$ is a set (or family) of probability distributions. These are paired with the edges of the graph.

---

References

@koski2011 - Chapter 2