Representation

• Each variable is represented by a node and has two types of information specified.
  1. Arcs representing the assertions that the variable is conditionally independent of its nondescendents given its immediate predecessors (i.e., Parents). is a descendent of if there is a directed path from to .
  2. A conditional probability table describing the probability distribution for that variable given the values of its immediate predecessors. This joint probability is computed by
• is conditionally independent of its nondescendents and given its parents and
• 
• Also notice that is conditionally independent of and given and and
• Similarly, is conditionally independent of , , , and given .
• BBNs are a convenient way to represent causal knowledge. The fact that causes is represented in the BBN by the fact that is conditionally independent of other variables in the network given the value of .