Question

You are given a dataset which has documents with class labels, for example {Document i, Sport} means that document i belongs Sport class. In the dataset, you calculate the total number of words in all Sport documents as 1000, and the total number of words in all Arts documents as 1000. Also, there are a total of 42 occurrences of w1, 54 occurrences of w2, and 33 occurrences of w3 in all Sports documents, and 58 occurrences of w1, 50 occurrences of w2, and 31 occurrences of w3 in all Arts documents. Also assume that the probability that a random document is a Sport document is 0.11 and the probability that it is an Arts document is 0.16, i.e. P(Class is Sports) = 0.11, P(Class is Arts) = 0.16. Given a document D = , use the Naive Bayes classifier and calculate the ratio of P( Sport |D) to P( Arts | D) ?

Answer 1