I’ve got invited to give a workshop about Bayesian Inference using Julia and Turing.jl at Stuttgart Julia Programming Language Meetup.
Saturday 12th June.
The workshop will be using a Reactive Pluto notebook. The code can be found on GitHub and the static notebook.
You can download it run on your machine or run in the cloud (using binder) by clicking on “edit or run this notebook” and them on “binder”.
Also note that the workshop are based on a set of Bayesian Statistics using Turing tutorials that can be found here: Bayesian Statistics using Julia and Turing
How to attend here: Bayesian Inference with Turing.jl | Meetup
There is a lot of attendees that would be way more experienced guide to Turing than me
I also just noticed that one of the people that I cite (Liza Semenova) in the Notebook will be attending also.
I’ve uploaded the video to YouTube with timestamps for the topics covered: https://youtu.be/CKSxxJ7RdAU
I just learned that a hidden Markov model is the name of the thing I was trying to reinvent. And now I know how to do it properly!
Just note that the marginalization with NUTS is much more efficient than the composite sampler that I’ve made. You can find the Stan model converted to Turing here: TuringExamples/turing.jl at master · TuringLang/TuringExamples · GitHub