What extra reference should I read?


#1

I’m reading tutorials, but there are some formulation I can understand completely. What book or blog should I read? Or just follow the tutorial, read more examples first?

In fact, I do not know much about probability. Before this, I do computer vision based on pytorch. So I need some advice…

Thanks!


#2

hi alex,

the tutorials are a great place to start - personally ive found hacking on a problem the best way to jump into a new area. the tutorials give you a good basic intro to models and inference, then walk you through examples of increasing complexity. you can also try taking an existing model youve written and try being bayesian about it ie do inference over some parameters of interest. good luck!


#3

some people seem to like

http://camdavidsonpilon.github.io/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/


#4

I think you can read some paper and tutorial in nips. The basic concept of pyro comes from bayes deep learning. I recommend: “Variational Inference Foundations and Modern Methods” and “Uncertainty in Deep Learning”.