Probabilistic modeling through Bayesian inference using PyStan with demonstrative case study experiments from Christopher Bishop's Model-based Machine Learning.

Python package realizing Abdul Malik's idea: "Wouldn't the idea behind the state_collapser package also give a speed-up for [insert favorite learning framework] by such a hierarchical version? A straightforward hierarchical version of Belief Propagation is what I see."

Working with sums of bivariate functions in Python