Conditional Probability Table (CPT)

Each discrete chance node in a Bayesian network or a LIMID is assigned a conditional probability table (abbreviated CPT). The node name and node label displayed in the Tables Pane when working in Edit Mode specifies a currently selected node. If this node is a discrete chance node, the table shown in the Tables Pane is a CPT.

Only if you have chosen to specify the table manually (alternative to using expressions ), you can see the entire CPT. Otherwise, you see the table of expressions specifying the table.

The table shown in the Tables Pane is different depending on the type of the currently selected node:

• If the currently selected node is a discrete chance node, the table is a true CPT.

• If the currently selected node is a utility node, the table is a utility table rather than a CPT.

• If the currently selected node is a decision node, the table represent the policies associated with the decision.

• If the currently selected node is a continuous chance node, the table specifies a Gaussian (normal) distribution function rather than a CPT.

Figure 1 shows a CPT for the discrete chance node Grass with two parents, Sprinkler and Rain.

Figure 1: The CPT for Grass.

The states of the currently selected node are those found in the left most column below the parent labels. You can change these state names by selecting them with the mouse cursor and typing in a new name from the keyboard.

Each numeric value in the CPT is the probability of the currently selected node being in the state found in the left most column in the actual row - when the parents (if any) are in the states found in the top of the actual column.

You can change a specific numeric value by selecting it with the mouse cursor and typing in a new value from the keyboard.