1. A Nave Bayes classifier is to be used to classify emails into Urgent (U) vs [2 MARKS] Non-urgent (N). The following 12 training examples are given to train the classifier. The features under consideration are: whether the emails are short, whether they contain known spam terms and whether they are received from a manager. (The Class column has been sorted for you.) c) It is observed that the contains some spam phrases. For each class, what is [14 MARKS IN TOTAL] the likelihood of this observation? [2 MARKS] d) It is observed that the email is not from a manager. For each class, what is the likelihood of this observation? [2 MARKS] e) What is the likelihood of all three of the above observations, b), c) and d), for both respective classes? [2 MARKS] a) For each class (U and N), what is the prior on the hypothesis? f) Would maximum a priori probability (MAP) or maximum likelihood (ML) be more suitable for this dataset? Why? [2 MARKS] [2 MARKS] b) It is observed that the email is long. For each class (U and N), what is the g) Given those observations, and your choice in f), which class is predicted, and likelihood of this observation? why? Activate Window' MARKS] Go to Settings to activate Windows