Eberhard Karls Universität Tübingen

Mathematisch-Naturwissenschaftliche FakultätProgramming Languages and Software Technology

Embedded Probabilistic Programming with Continuous Variable Distributions

Probabilistic programming languages are programming languages which have special support to describe probabilistic models, and then perform inference in these models. They are attractive because they unify the expressive power of a general-purpose programming language with probabilistic modeling and reasoning.

A particularly elegant way to implement probabilistic programming is to embed it into an existing general-purpose language by means of libraries. The HANSEI library, for instance, adds sophisticated probabilistic programming to OCaml.

Existing embedded probabilistic programming languages support only discrete probability distributions. The purpose of this thesis is to extend HANSEI with support for continuous probability distributions and appropriate inference methods, such as Markov Chain Monte Carlo sampling.

This thesis is well-suisted for a student with a good background in functional programming and some background in statistics.

Further Information

Contact

Klaus Ostermann