Inference in Bayesian networks Consider the alarm Bayesian network. Each of these random variables is binary-valued. F indicates whether your home is on fire, E indicates whether an earthquake is happening, A indicates whether your alarm system has gone off, B indicates whether your neighbor Bob has called to tell you that your alarm has sounded, and C indicates whether your neighbor Carol has called to tell you that your alarm has sounded. The local conditional probability distributions for each of these random variables are as follows: P(F=1) = 0.01 P(E=1) = 0.02 f e P(A=1| F=f, E=e) 1 1 0.95 1 0 0.94 0 1 0.29 0 0 0.01 a P(B=1| A=a) 1 0.90 0 0.05 a P(C=1| A=a) 1 0.70 0 0.01 Perform exact inference by enumeration to compute P(F = 1 | B = 1)