Data-driven Sequential Monte Carlo in Probabilistic Programming

TL;DR

This paper introduces a neural network-based approach to improve proposal distributions in Sequential Monte Carlo algorithms within probabilistic programming, leading to more efficient inference with fewer particles.

Contribution

It presents a method to train neural networks to approximate optimal proposals using posterior estimates, enhancing SMC inference in probabilistic programming systems.

Findings

Data-driven proposals significantly improve inference efficiency.

Fewer particles are needed for accurate posterior estimation.

Applied in Anglican probabilistic programming system.

Abstract

Most of Markov Chain Monte Carlo (MCMC) and sequential Monte Carlo (SMC) algorithms in existing probabilistic programming systems suboptimally use only model priors as proposal distributions. In this work, we describe an approach for training a discriminative model, namely a neural network, in order to approximate the optimal proposal by using posterior estimates from previous runs of inference. We show an example that incorporates a data-driven proposal for use in a non-parametric model in the Anglican probabilistic programming system. Our results show that data-driven proposals can significantly improve inference performance so that considerably fewer particles are necessary to perform a good posterior estimation.