Also read Rasmunnsen.
Notes
- Regression with custom priors : Based on an example from the pymc3 docs and Jake Vanderplas's must-read blog article at http://jakevdp.github.io/blog/2014/03/11/frequentism-and-bayesianism-a-practical-intro/ , we do in pymc a custom regression with uninformative priors. We cover how to create custom densities and how to use theano shared variables to get posterior predictives.
- GP Recap and example :