Hi
Suppose there are three nodes, A, B, c
A>c B>c that is, A and B are the parent nodes of the cnode.
We know that P (A), P (B), P (c  A), P (c  B)，How do I get the conditional probability table for c when I set up a Bayesian network in GeNie?
If the parent node is too much, CPT also need to enter one by one?
How do I get a conditional probability table?

 Site Admin
 Posts: 299
 Joined: Tue Dec 11, 2007 4:24 pm
Re: How do I get a conditional probability table?
I'm afraid no Bayesian network software will allow you to do that  you have a theoretical problem. You need P(CAB) and have only P(CA) and P(CB). Without special assumptions about the interaction, you are facing an underconstrained problem.
Please look at the canonical models (Section "Components of GeNIe models" and then "Canonical models"). Even there, though, you need to have P(CA~B) and P(C~AB).
I hope this helps,
Marek
Please look at the canonical models (Section "Components of GeNIe models" and then "Canonical models"). Even there, though, you need to have P(CA~B) and P(C~AB).
I hope this helps,
Marek