Nodes not having an effect on sensitivity

[User2025]

I created a BN with three sets of influence chains all influencing a final target node. However, when running a strength of influence analysis, one of the chains has no effect on the final node. Its sensitivity is 0, while every other node has some strength of influence. How can I fix this?

[marek [BayesFusion]]

Thank you for sending me your model. I have looked at it and saw that quite likely you mean the nodes n0, n6, n3, n4, n12 , n11 and n9 in your obfuscated model. Each of these nodes is defined by a CPT that consist of only zeros and ones. Zero and one are special numbers in probability theory. Once a probability is zero, it can never change, no matter how strong evidence is. Consider Bayes theorem P(A|B)=P(B|A)*P(A)/P(B). If P(A)=0, P(A|B) has to be zero, no matter how strong the evidence B. Those nodes that have probability zero (or one) cannot ever change their posterior probabilities and, hence, will never play any role in sensitivity analysis. Does this help?
Cheers,

Marek

[User2025]

Hi,

I was specifically interested in the strength of influence parameter. Only nodes 8 and 9 re zeroed out. Everything else, even other nodes with 0/1 probabilities, have positive influence.

[marek [BayesFusion]]

The guilty element of your model is the CPT in n10:

s0 0.5 0.5 0.5 0.5
s1 0.5 0.5 0.5 0.5

Please note that the probability distribution over n10 is not impacted by the parents.
Does this help?

Marek

[User2025]

how would I make it impacted by the parents?

[marek [BayesFusion]]

By changing the definition of the child node. The conditional probability distributions describe precisely the impact that the parents have on it. In your model, the conditional probability distributions (columns in your CPT) are all equal.

I hope this helps,

Marek