Semi-Symbolic Inference for Efficient Streaming Probabilistic Programming

TL;DR

This paper introduces semi-symbolic inference, a method that automatically combines exact and approximate inference in streaming probabilistic programming, enabling efficient Rao-Blackwellized particle filtering with significant speedups on certain benchmarks.

Contribution

The work presents a novel semi-symbolic inference technique that automates Rao-Blackwellized particle filtering in probabilistic programming, leveraging symbolic distributions for exact inference when possible.

Findings

Achieves 1.6x slowdown on existing benchmarks

Attains 3x to 87x speedups on new challenging benchmarks

Automatically implements RBPF similar to hand-coded solutions

Abstract

Efficient inference is often possible in a streaming context using Rao-Blackwellized particle filters (RBPFs), which exactly solve inference problems when possible and fall back on sampling approximations when necessary. While RBPFs can be implemented by hand to provide efficient inference, the goal of streaming probabilistic programming is to automatically generate such efficient inference implementations given input probabilistic programs. In this work, we propose semi-symbolic inference, a technique for executing probabilistic programs using a runtime inference system that automatically implements Rao-Blackwellized particle filtering. To perform exact and approximate inference together, the semi-symbolic inference system manipulates symbolic distributions to perform exact inference when possible and falls back on approximate sampling when necessary. This approach enables the system to implement the same RBPF a developer would write by hand. To ensure this, we identify closed families of distributions -- such as linear-Gaussian and finite discrete models -- on which the inference system guarantees exact inference. We have implemented the runtime inference system in the ProbZelus streaming probabilistic programming language. Despite an average $1.6\times$ slowdown compared to the state of the art on existing benchmarks, our evaluation shows that speedups of $3\times$-$87\times$ are obtainable on a new set of challenging benchmarks we have designed to exploit closed families.