I want to understand the significance of a posterior probability in a Bayesian Network (BN)? I am confused regarding its application and significance.
In my case, I have a power pant with equipment failure data with me. With prior probabilities and CPT in a BN, I have obtained the probability of failure of the generating unit of the plant.
Now I wish to know, if I put the failure of the generating unit as 100%, I will obtain the posterior probabilities of all the above parent, child and root nodes. This means if I observe unit failure, what is the contribution of nodes in the BN?
But I do not understand the logic of doing this, since the failure data has not changed, hence prior and CPT has not changed? Is it giving the impact or significance of each node towards the resulting event of generating unit failure?
Is it the same as data updating in a BN?
Thank You!