Trying something a little bit different today: what subjects would you like to see me write about? Any topics of particular interest? Any posts that you’d like me to go further on?
As a beginner Bayesian stats practitioner, I’d love to see examples of how you take a business problem, translate it to a specific approach, and move on to recommendation. Few can write about this stuff as clearly as you can.
I'd be interested in a series of like "How to STAN" type posts (e.g., what is it, how to setup your environment to do STAN work, building a model, validating it, testing it, etc etc)
How do you build your causal models? I suppose it means building causal DAGs beforehand, but then how are those designed? Do you use the data to build (or invalidate) some relations? How much energy is spent during this designing process? How much DAGs come out of it that need to be investigated?