DAG or Bayesian Network helps us to reduce the number of parameters to be stored drastically.
Store only factored joint probabilities instead of joint probabilities of all the variables.
Dependencies
Child node depends on Parent node, but is independent of Grand Parent node
Child node is independent of sibling node given a Parent node
Parent nodes are NOT independent given a Child node.
Markov Blankets
The Markov Blanket of a node (Xi) is the set that renders it independent of the rest of the graph. This is the parents (U), children (Y), and co-parents of children (Z).