SIXO: Smoothing Inference with Twisted Objectives

TL;DR

SIXO introduces a novel approach to inference in state space models by learning targets that approximate smoothing distributions, leveraging density ratio estimation to improve accuracy and theoretical bounds in posterior inference and model learning.

Contribution

The paper presents SIXO, a new method that uses density ratio estimation to learn targets approximating smoothing distributions, enhancing inference accuracy and theoretical guarantees.

Findings

SIXO provides tighter log marginal lower bounds.

SIXO yields more accurate posterior inferences.

SIXO improves parameter estimation across domains.

Abstract

Sequential Monte Carlo (SMC) is an inference algorithm for state space models that approximates the posterior by sampling from a sequence of target distributions. The target distributions are often chosen to be the filtering distributions, but these ignore information from future observations, leading to practical and theoretical limitations in inference and model learning. We introduce SIXO, a method that instead learns targets that approximate the smoothing distributions, incorporating information from all observations. The key idea is to use density ratio estimation to fit functions that warp the filtering distributions into the smoothing distributions. We then use SMC with these learned targets to define a variational objective for model and proposal learning. SIXO yields provably tighter log marginal lower bounds and offers significantly more accurate posterior inferences and parameter estimates in a variety of domains.