Question 1. (100 points) Consider the following Bayesian belief network (BBN). Part a. How many probability numbers are needed for the full joint probability table? How many are needed for the above BBN? Consider the following query for parts b through e : P(cb,g) Part b. Compute the above probability using the variable enumeration algorithm. This is the formula based approach. Clearly show your steps. Part c. Show how two full samples could be generated using rejection sampling. The first sample should be one that gets rejected, and the second one should be one that does not get rejected. Explain how the probability above would be calculated. Part d. Show how a single full sample could be generated using likelihood weighting (LW). Explain how the probability above would be calculated. Part e. Show the sample space for the Markov Chain Monte Carlo (MCMC) algorithm. You are not required to show the sample space pictorially. You can list as text. Show how the initial sample and a single full sample other than the initial sample could be generated using MCMC. Explain how the probability above would be calculated