CLOSE REVIEWWhat is a prior probability?
(Student response here)
Did you like this question?
Instructor solution
A prior probability, or unconditional probability, is the initial likelihood of an event occurring given no additional information. For instance, with a deck of cards, our prior probability of drawing an ace would be a 1/13. Prior probabilities are used in Naive Bayes to calculate the likelihood of an input mapping to each possible output, such as the probability that Document 1 belongs to Label A is equal to the prior probability of any document belonging to Label A multiplied by the probabilities of each word in Document 1 belonging to Label A (normalized).
Was this helpful?
Speech and Language Processing: Naive Bayes and Sentiment Classification (page 4)
Was this helpful?
Think you've got it?
For a binary sentiment analysis task (labels: - and +), the prior probability for + is _.
- A.
not possible to determine since there are an infinite amount of potential examples
- B.
the count of training examples for + divided by the total count of training examples
- C.
larger than the prior probability for -
- D.
1/2
Was this helpful?
You may exit out of this review and return later without penalty.