Exercise 60 A psychologist does a small survey on 'happiness'. Each respondent provides a vector with entries I or 0 corresponding to whether answer "yes' to a question or 'no', respectively. The question vector has attributes I = (rich, married, healthy) In addition, each respondent gives a value c = 1 if they are content with their lifestyle, and c = 0 if they are not. Thus, a response (1, 0. 1) would indicate that the respondent was 'rich' 'unmar- ried', 'healthy'. The following responses were obtained from people who claimed also to be 'content' (1, 1, 1), (0, 0, 1), (1, 1, 0), (1, 0, 1) For the 'not content' respondents, the data is (0, 0, 0), (1, 0, 0), (0, 0, 1), (0, 1, 0), (0, 0, 0) Using Naive Bayes on this date, what is the probability that a person who is 'not rich', 'married' and 'healthy' is 'content'? What is the probability that a person who is 'not rich' and 'married' is 'content'? (That is, we do not know whether or not they are healthy'). Consider the following vector of attributes : 21 = 1 if customer is younger than 20 ; 21 = 0 otherwise I2 = 1 if customer is between 20 and 30 years old : 22 = 0 otherwise I3 = 1 if customer is older than 30 ; 23 = 0 otherwise IA = 1 of customer walks to work : IA = 0 otherwise Each vector of attributes has an associated class label "rich" or "poor". Point out any potential difficulties with using your previously described approach to training using Naive Bayes. Hence describe how to extend your previous Naive Bayes method to deal with this dataset. Describe in detail how marimum likelihood could be used to train this model