Building some Data analysis Tutorials

So, for those who are interested in following along. Like many things these days, I’ve been derailed a bit by COVID. Specifically I have a lot of friends and family who want to know what the latest info is on the COVID epidemic, and I was making some PDFs by hand and putting them on my blog each week or so, but I figured, hey, why not give them julia notebooks they can interact with… And I managed to get that into binder etc, It’s been educational, but it’s a work in progress and not very tutorial-like really. In particular I don’t have a discussion document for the COVID stuff because it’s still a work in progress.

I’d like to do a tutorial in which I use Turing to build a Bayesian model of something interesting. Here’s your chance to influence that. What would you like to see modeled? Requirements are:

  1. Publicly available dataset, prefer something not too enormous. Must be an easily readable format (CSV for example). It could involve integrating data from two public sources.
  2. Model shouldn’t require tons of moving parts (so for example the COVID epidemic while very interesting, is a very challenging field, so it’s out). Also shouldn’t be trivially simple (something you could do fine with GLM and a linear or logistic regression with a point estimate).
  3. Should be a topic I have some familiarity with: Economics, Biology, Healthcare, Mechanics/Physics, Civil and Environmental Engineering would be good candidates.

Thoughts?

1 Like