Variance estimation for Sequential Monte Carlo Algorithms: a backward sampling approach - Université de Paris - Faculté des Sciences Access content directly
Preprints, Working Papers, ... Year : 2022

Variance estimation for Sequential Monte Carlo Algorithms: a backward sampling approach

Abstract

In this paper, we consider the problem of online asymptotic variance estimation for particle filtering and smoothing. Current solutions for the particle filter rely on the particle genealogy and are either unstable or hard to tune in practice. We propose to mitigate these limitations by introducing a new estimator of the asymptotic variance based on the so called backward weights. The resulting estimator is weakly consistent and trades computational cost for more stability and reduced variance. We also propose a more computationally efficient estimator inspired by the PaRIS algorithm of [33]. As an application, particle smoothing is considered and an estimator of the asymptotic variance of the Forward Filtering Backward Smoothing estimator applied to additive functionals is provided.
Fichier principal
Vignette du fichier
arxivpaper.pdf (978.39 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03630333 , version 1 (05-04-2022)
hal-03630333 , version 2 (10-01-2023)

Identifiers

  • HAL Id : hal-03630333 , version 1

Cite

Yazid Janati, Sylvain Le Corff, Yohan Petetin. Variance estimation for Sequential Monte Carlo Algorithms: a backward sampling approach. 2022. ⟨hal-03630333v1⟩
108 View
96 Download

Share

Gmail Mastodon Facebook X LinkedIn More