A directed acyclic graph is a directed graph with no path returning to its starting node. It represents relationships that flow between variables without forming cycles. Bayesian networks use this structure to map probabilistic dependencies and show how changes in one event can affect others. section 3.5
Weather prediction
A weather network can connect a clouds node to a rain node with an arrow from clouds to rain. If clouds have a 40% likelihood, the model applies different rain probabilities for cloudy and non-cloudy conditions to estimate possible scenarios. section 3.5
Figure 3.5 An example of a Bayesian network
Use the network figure to inspect how variables and directed dependencies form a Bayesian model. section 3.5
