50+ days of Machine Learning

 Hello :) Today is Day 125! A quick summary of today: listened to day 2 lectures on causality and optimization of MLx Fundamentals learned a bit of STATA The schedule was: The recording of the lectures was released about 30 mins ago, so going over it will be my task for tomorrow.  The 1st lecture from Professor Kun Zhang from CMU was more specifically about causal representation learning.  For example finding hidden variables.  Here if we look at just the relationship between cholesterol and exercise (right graph) we can see they have a positive relationship. Which is quite weird, and when we incorporate age into the picture, we can see the actual negative relationship.  In this case there are treatment A and B for kidney stones, and if we just look at the overall, without accounting for stone size, we might conclude that B is better. But if we incorporate stone size into the picture, A is better in both cases. if we understand the problem well, there is no paradox because we know th