Learning Probabilistic Programs
This work addresses the challenge of scalable and reusable inference in probabilistic programming for machine learning practitioners, though it appears incremental in building on existing techniques.
The paper tackles the problem of generalizing from data using probabilistic programs as samplers, showing that Markov chain Monte Carlo inference with higher-order probabilistic programming languages enables successful inference in nontrivial domains. It also introduces a new notion of probabilistic program compilation for efficient reusable predictive inference.
We develop a technique for generalising from data in which models are samplers represented as program text. We establish encouraging empirical results that suggest that Markov chain Monte Carlo probabilistic programming inference techniques coupled with higher-order probabilistic programming languages are now sufficiently powerful to enable successful inference of this kind in nontrivial domains. We also introduce a new notion of probabilistic program compilation and show how the same machinery might be used in the future to compile probabilistic programs for efficient reusable predictive inference.