Issue on page /21-Meta-Learners.html Small correction on the X-learner explanation #454
Description
Hi Matheus, I believe there’s a small inversion in this part of the explanation:
More generally, weighted average using the propensity score will make sure we give more weight to the CATE model that was estimated where the assigned treatment was more likely.
If the propensity of a user to belong in treatment is high, say 0.9. Then according to the formula, the weight for the treatment model will be 0.1 and 0.9 for the control model. So the statement is in reverse.
Thank you for making this book available for free! Im really enjoying it.