Automated Variational Inference in Probabilistic Programming
This addresses the challenge of efficient inference for complex, non-tractable distributions in probabilistic programming, which is incremental.
The authors tackled the problem of approximate inference in probabilistic programs by developing a new algorithm based on stochastic gradient for variational programs, which improves inference efficiency over other algorithms without restrictions on the probabilistic program.
We present a new algorithm for approximate inference in probabilistic programs, based on a stochastic gradient for variational programs. This method is efficient without restrictions on the probabilistic program; it is particularly practical for distributions which are not analytically tractable, including highly structured distributions that arise in probabilistic programs. We show how to automatically derive mean-field probabilistic programs and optimize them, and demonstrate that our perspective improves inference efficiency over other algorithms.