I can see how this impacts predictive algorithms. Once you know that "department" is a crucial variable, you go back to the prediction and make sure it is included in the training dataset, so that the prediction will be made not only according to "gender", but to "department" as well.
I think Granger is more of a traditional econometrics topic? Some mentioned that it’s not exactly in the same formulation of causal inf procedure. But I’ll have to read more
Oh after a closer look it does not cover some of the preprocessing but included some acyclic graph computation. It has ATE and propensity stuff and more on orthogonal decomp of models
I can see how this impacts predictive algorithms. Once you know that "department" is a crucial variable, you go back to the prediction and make sure it is included in the training dataset, so that the prediction will be made not only according to "gender", but to "department" as well.
beautiful talk. loved the fast pace
讲的快但是非常凝练,很牛逼!Awesome
Have you discussed Granger causality or did I miss something?
I think Granger is more of a traditional econometrics topic? Some mentioned that it’s not exactly in the same formulation of causal inf procedure. But I’ll have to read more
Hi, It will be nice to have the code or the notebook that you used. Thank you so much, please let know where we can download from.
There’s a book causal inference in Python on o reily that closely follows this. I saw an open source book as well but forgot it’s name
Oh after a closer look it does not cover some of the preprocessing but included some acyclic graph computation. It has ATE and propensity stuff and more on orthogonal decomp of models