Naïve Bayes using the Binomial (also known as Bernoulli) model. First calculate estimates of P(c)
and P(w|c) given the following 10 training sentences:
Training data
1 Travel: chicago mayor takes vacation in hawaii
2 Travel: nurses plan a trip to hawaii
3 Travel: employers offering more jobs with vacation benefits
4 Business: employers see growth in computer sector
5 Business: high paying retail jobs in chicago
6 Business: nurses find employers spending more on hospital jobs
7 Business: nice vacation spot but no jobs in hawaii
8 Health: nurses need to take a vacation
9 Health: doctors attend hawaii conference
10 Health: chicago employers have jobs for nurses
There are only three classes: Travel, Business, and
Health. You should not use any smoothing. For features you should only use the following six vocabulary
terms: {chicago, employers, hawaii, jobs, nurses, vacation} and you should ignore all other words and any
punctuation. Next compute the predicted class for the two test documents A & B below. Please read these
directions thoroughly, count carefully, and do show all of your work.
Test documents:
A: nurses take golf vacation in hawaii
B: top employers move jobs to chicago
0 comments