LFCS Seminar: Tuesday, 13 June - Mathieu Huot

Title:       ADEV: Sound Automatic Differentiation of Expected Values of Probabilistic Programs

Abstract:

Optimizing the expected values of probabilistic processes is a central problem in computer science and its applications, arising in fields ranging from artificial intelligence to operations research to statistical computing. Unfortunately, automatic differentiation techniques developed for deterministic programs do not in general compute the correct gradients needed for widely used solutions relying on gradient-based optimization.

I will talk about ADEV, an extension to forward-mode automatic differentiation that correctly differentiates the expectations of probabilistic processes represented as programs that make random choices. Our algorithm is a source-to-source program transformation on an expressive, higher-order language for probabilistic computation, with both discrete and continuous probability distributions. The result of our transformation is a new probabilistic program, whose expected return value is the derivative of the original program’s expectation. This output program can be run to generate unbiased estimates of the desired gradient. As it modularly extends forward-mode AD, our algorithm lends itself to a concise implementation strategy, which we exploit to develop a prototype in just a few dozen lines of Haskell (https://github.com/probcomp/adev).

Themes: functional programming, semantics, logical relations, automatic differentiation, probabilistic programming.

Short Bio: Mathieu Huot is finishing his PhD at the University of Oxford under the supervision of Pr. Sam Staton. He will be joining the ProbComp team at MIT as a research scientist. His research is centred around the semantics of programming languages, category theory, automatic differentiation, and probabilistic programming.