15

15 Graphical Models w_good h_true h_false w_bad 0.4 0.3 Weather Holiday Ice Cream Sales #Tweets s_high s_low #low #high 0.8 h_true h_false 0.9 w_bad, h_true w_bad, h_false w_good, h_true w_good, h_false 0.2 0.4 A Bayesian Network Please fill in the following blanks. 4.0p a To explicitly define a joint distribution over four random variables: Weather (w_good or w_bad), Holiday (h_true or h_false), Ice Cream Sales (s_high or s_low), and #Tweets (#high or #low), the number of probabilities to specify is at least given no independence assumptions. 3.0p b Given the above Bayesian network, we only need to specify conditional probabilities. 5.Op The probability of (good weather A holiday A high ice cream sales 1 a high number of tweets) is Please decide whether the following statements are True or False. 1.Op d According to the above Bayesian network, given high ice cream sales, Weather and Holiday are independent. True False 1.Op e When a variable's parents are observed in a Bayesian network, this variable is independent of all the other variables. True False 1.0p f One single node can be selected as a clique of a Markov random field when computing joint probabilities. True False