You are given three different Bayesian network structures (1, 2, 3), for three alarm system designs. Each system consists of 4 binary random variables (A, B, C, D). Each variable corresponds to an alarm condition in an assembly line in a factory, whose state can be either “ON” or “OFF”. Structures in image Using the chain rule of probability for Bayes Nets, factor P(A,B,C,D) according to the independencies specified in the three structures above. What is the minimum number of parameters required to fully specify the distribution (i.e. to calculate P(A,B,C,D))? Hint: The chain rule of probability P(A,B) = p(A|B) p(B) for three variables: P(A,B,C) = P(A| B,C) P(B,C) = P(A|B,C) P(B|C) P(C) and in general to n variables: P(A1, A2, ..., An) = P(A1| A2, ..., An) P(A2| A3, ..., An) P(An-1|An) P(An)
You are given three different Bayesian network structures (1, 2, 3), for three alarm
Structures in image
Using the chain rule of probability for Bayes Nets, factor P(A,B,C,D) according to the independencies specified in the three structures above. What is the minimum number of parameters required to fully specify the distribution (i.e. to calculate P(A,B,C,D))?
Hint: The chain rule of probability
-
P(A,B) = p(A|B) p(B)
for three variables:
-
P(A,B,C) = P(A| B,C) P(B,C) = P(A|B,C) P(B|C) P(C)
and in general to n variables:
-
P(A1, A2, ..., An) = P(A1| A2, ..., An) P(A2| A3, ..., An) P(An-1|An) P(An)
Trending now
This is a popular solution!
Step by step
Solved in 3 steps