We’re building a model to predict spam email. The positive cases are spam and the negative cases are legitimate. If we’re going

Question

here the answer is precision. likewise, can you provide a situation with recall ?

Answer 1

Ok, for your example, if we decide to delete emails that are spam then we should be extremely careful not to delete important emails thinking that it is spam, so we should minimize false positives i.e increase Precision. However, if we don't decide to delete spams then no need to minimize false positives which will increase the recall✨.