Even strong classifiers can suffer low precision when the positive class is rare. I walk through some simple math to show what precision can look like with low base rates.
What is a prediction interval? How would you use one?
I go over the most common interpretation of a confidence interval, and discuss an alternative, potentially more useful one.
I go over OLS regression in matrix notation, and show how it connects to the scalar notation of OLS regression.
A “warm up” post covering simple ordinary least squares regression.
