could you help me on this problem ? Thank you .

Consider the following training data: The goal is to predict whether new subjects with given bu,s_computer s 30 high no fair 5 30 ' excellent no [31,40] ' fair yes > 40 fair yes > 40 fair yes > 40 excellent no [31,40] excellent yes 5 30 fair yes 5 30 fair yes > 40 fair yes 5 30 excellent yes [31,40] excellent yes [31,40] ' fair yes > 40 excellent no features (age, income, student, credit rating) would buy a computer, using naive Bayes classier. (i) Compute the maximum likelihood prior distribution on the label {0,1}, where 0 ='does not buy com- puter' and 1 ='buys computer'. (ii) We will model the feature space as the following lO-dimensional product space of binary values {0. 1}:1 (pun) = [1(age s 30), 1(age e [31,40]), 1(age > 40),..., 1(credit = ecellent)] e {0,1}1. (5) Compute the maximum likelihood class-conditional probabilities of each feature. (iii) Suppose we have the following testing examples: x1 = \" age 5 30, medium income, student, fair credit rating\" (6) x2 = " age 6 [3 1, 40], low income, not student, low credit rating\" (7) x3 = " age > 40, high income, not student, excellent credit rating\" (8) Compute the predictive probabilities {posterior distribution of class labels} of each testing exam- ples. Make the prediction. For each example, can you tell which factor affected the most for the Plneuir'nnn rpm Ilf