Bayesian ML and Probabilistic Programming

When the understanding of phenomena and interpretation of results are required in industrial or scientific applications, black-box models give way to hand-crafted (or automatically discovered) probabilistic models. Classical Bayesian methods do not scale well to high dimensions or large datasets and are plagued by intractable computations. But with the development of expressive and elegant probabilistic programming languages like Stan or Pyro and their incorporation of advanced approximate inference methods, today’s practitioners benefit from a powerful arsenal that enables faster development of more complex models than ever before.