I am reading ISLR and am on page 132 on chapter 4. The author mentions the term "average fitted probability". What is that?
So if we divide the number of people who default by the total number of people in our training set (marginal probability), we get 0.0333. And that is related to the logistic regression in what way?