Bayes nets

Question 1

Why bayes nets are important?

Answer 1

• Takes probability and uncertainty and marries them with efficient structures
• It shows how a uncertain variable influences another uncertain variable

Question 2

Bayes rule

Answer 2

• P(A|B) = P(B|A)*P(A)/P(B)
• P(-A|B) = P(B|-A)*P(-A)/P(B)
• P’(A|B) = P(B|A)*P(A)
• P’(-A|B) = P(B|-A)*P(-A)
• n = (P(B|A)P(A) + P(B|-A)P(-A))^-1

Question 3

Conditional independence

Answer 3

• P(T2|C,T1) = P(T2|C)
• B and C are conditionally independent given A (If we know A). If we do not know A then B and C are not conditionally independent

Question 4

Explaining way

Answer 4

• Example: sunny&raquo_space; happy, raise&raquo_space; happy

If I am happy and it is sunny, then it is less likely that I received a raise

Question 5

What is a bayes network?

Answer 5

• It defines the probability distribution over graphs or random variables
• It requires less parameters than a full probability distribution. It is more compact and efficient.

Question 6

What is D-Separation?

Answer 6

• Two variables are independent if not link by unknown variables. An exception is when the “explain way effect” happens
• It is the situation in which two node in a bayes net become independent or dependent