Question

Now let’s assume that we observe the following three emails with their true label in parentheses:

(Spam): This is a warning that your social security number has been stolen.
(Ham): More cat and dog photos?
(Ham): I love exam prep, regular sections and social events.

What are the estimates of the following probabilities using the Naive Bayes model?
- P (W = warning| = spam) A. 0 B. 1∕10 C. 1∕5 D. 1∕3 E. 2∕3 F. None of the options
- P (W = social| = ham) A. 0 B. 1∕10 C. 1∕5 D. 1∕3 E. 2∕3 F. None of the options
- P (W = office| = ham) A. 0 B. 1∕10 C. 1∕5 D. 1∕3 E. 2∕3 F. None of the options
- P ( = ham) A. 0 B. 1∕10 C. 1∕5 D. 1∕3 E. 2∕3 F. None of the options

Answer 1

Using the Naive Bayes model, we can estimate the probabilities of certain words occurring in spam and ham emails.

Step-by-step explanation:

Using the Naive Bayes model, we can estimate the probabilities as follows:

1. P (W = warning | spam) = 1/3
2. P (W = social | ham) = 1/5
3. P (W = office | ham) = 0
4. P (ham) = 2/3

The estimate for the probability of 'warning' given that the email is spam is 1/3. The estimate for the probability of 'social' given that the email is ham is 1/5. There is no occurrence of 'office' in the ham emails. The estimate for the probability of an email being ham is 2/3.