I have almost finished studying 'Causal Inference in Statistics: A Primer', but I still feel that I need to learn more.
I considered 'Causality' (Pearl, 2009), but there seem to be several good learning resources about DAG (ex. Review Paper & etc).
What should I study after finishing 'Causal Inference in Statistics: A Primer'?
$\begingroup$
$\endgroup$
8
-
4$\begingroup$ Cunningham's The Mixtape book or Huntington-Klein's The Effect book would be suitable suggestions for a new starter. (Both books are free online too.) For something more formal would be the What If book from Hernan & Robins and Elements of causal inference by Peters et al. These four freely available books would most likely cover the vast majority of CI use cases one come across. $\endgroup$– usεr11852Commented May 28, 2022 at 11:56
-
$\begingroup$ @usεr11852 Thank you for answering my question. I interest 'Directed Acyclic Graph' for data analysis in Web Platform Business(ex. SNS, Commerce etc). I've been interested in the book 'Elements of causal inference' ever since. Can you give me a rough idea of what the 'Elements of causal inference' book contains? $\endgroup$– vinsh_77Commented May 28, 2022 at 12:14
-
1$\begingroup$ It is the most advanced of the four books mentioned so I am not sure it is very constructive without enough background. (I am also not a huge fan of having very few exercises.) Dr. Peters has some excellent YouTube content, for example, the series here - probably go through those first and refer to the book later. I know it is "uncool", but if you are interested in business applications do look at Econometrics literature. Pearl's work is not in a vacuum. That group has a lot of causal inference mileage. $\endgroup$– usεr11852Commented May 28, 2022 at 13:06
-
1$\begingroup$ (If you want to do "proper" DAGs, do check Probabilistic Graphical Models by Koeller and Friedman. It is exceptional for the mathematical/theoretical background.) $\endgroup$– usεr11852Commented May 28, 2022 at 13:11
-
1$\begingroup$ Sure, hit it. As mentioned probably do the lecture first, I found them very informative. 1. Regarding "proper": apologies, I meant it instead of "really good", it has informal use like: "Kevin is a proper player". I meant it provides a clear and rigorous foundation for the use of Bayesian Networks. 2. Pff... For PGM probably a class in Probability/Mathematical Statistics already. It is not meant to be an intro and neither is the Elements of Causal Inference. $\endgroup$– usεr11852Commented May 28, 2022 at 15:29
|
Show 3 more comments