On backward smoothing algorithms

TL;DR

This paper introduces a new framework for backward smoothing algorithms in state-space models, achieving linear complexity and providing stability guarantees, addressing limitations of existing methods.

Contribution

It presents a unified framework for skeleton-based smoothing algorithms, proves a novel stability theorem, and introduces a new coupling-based algorithm for intractable models.

Findings

Achieves truly linear complexity in smoothing algorithms.

Provides a non-asymptotic stability theorem.

Demonstrates practical effectiveness through numerical experiments.

Abstract

In the context of state-space models, skeleton-based smoothing algorithms rely on a backward sampling step which by default has a $\mathcal O(N^2)$ complexity (where $N$ is the number of particles). Existing improvements in the literature are unsatisfactory: a popular rejection sampling -- based approach, as we shall show, might lead to badly behaved execution time; another rejection sampler with stopping lacks complexity analysis; yet another MCMC-inspired algorithm comes with no stability guarantee. We provide several results that close these gaps. In particular, we prove a novel non-asymptotic stability theorem, thus enabling smoothing with truly linear complexity and adequate theoretical justification. We propose a general framework which unites most skeleton-based smoothing algorithms in the literature and allows to simultaneously prove their convergence and stability, both in online and offline contexts. Furthermore, we derive, as a special case of that framework, a new coupling-based smoothing algorithm applicable to models with intractable transition densities. We elaborate practical recommendations and confirm those with numerical experiments.