Synthetic Nodes

The size of a CPD table

In a Bayesian network, the size of the CPD table for a child node depends on the number of parent nodes.

For $N$ parent nodes, each having $k$ possible values, there will be $k^{N}$different combinations. The child node itself can also have $m$ possible values. Therefore, $m$ represents the number of possible values (or states) for the child node. Just as $k$ ...