I’ve been recently playing around with ‘arviz’. For those of you who don’t know Arviz is a library for exploratory analysis Bayesian Models.
I’ve got a Bayesian model built – or someone has built one for me, how do I explore it? How do I plot it?
This is the fundamental question that Arviz answers.
In this notebook I take a model I built, and get several interesting plots out of the model. For example it shows that Ireland is the best team at not conceding points. Defs should be negative, and the higher the better. This sorts of insights correlate with our domain specific understanding of the sport. We also see the uncertainty on each of these estimates – which is tremendously powerful.
Want to learn more?
If you’d like to learn more about Bayesian Analysis, or Probabilistic Programming – then I’ve put together an awesome course on this. It contains over 2 hours of exclusive content based upon me teaching these topics to over 400 scientists and engineers. I’ll be adding more content over the period. So if you’re interested you can sign up here.