Metraipolis › archive › dc_fd04699779bf
Metropolized backward sampling for conditional sequential Monte Carlo — arXiv:2610.38753
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.38753
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-01T02:30:56.147Z
- SHA-256 of the archived text
bac04a6269c03c498eb1a3872b115c5f6b4e284fde313ab0c7213f67ceb16eda- Stable URL of this page
https://metraipolis.com/documento/dc_fd04699779bf
How to cite
Metropolized backward sampling for conditional sequential Monte Carlo — arXiv:2610.38753 — arXiv (en), https://arxiv.org/abs/2610.38753 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fd04699779bf (Metraipolis archive dc_fd04699779bf, sha256 bac04a6269c03c498eb1a3872b115c5f6b4e284fde313ab0c7213f67ceb16eda)
BibTeX
@misc{dcfd04699779bf,
title = {Metropolized backward sampling for conditional sequential Monte Carlo — arXiv:2610.38753},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.38753},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fd04699779bf; sha256 bac04a6269c03c498eb1a3872b115c5f6b4e284fde313ab0c7213f67ceb16eda},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.38753. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Particle Gibbs samplers with backward sampling are widely used for sampling the smoothing distribution of state-space models, but, with few particles, the new trajectory often coincides with the current one at many time steps. A known but rarely used remedy replaces the random selection of the final particle by a Metropolis--Hastings move which avoids the current particle. We show that, with backward sampling, this remedy only helps at the last time steps, because its effect on the update rates decays geometrically with the distance to the final time. We propose instead Metropolized backward sampling, which makes such a move at every step of the backward pass, forcibly switching trajectories whenever possible. Additionally, we introduce a coupled version which prevents proposing the current trajectory, maximizing the chances of an actual switch versus the current trajectory and performing perfect sampling when the state-space model has no observations, a feature not shared by backward sampling. We prove that, at every time step, both methods dominate backward sampling as they update the state more often and have smaller lag-one autocorrelations than it does. Our methods are drop-in replacements for backward sampling, and can be used to improve any particle Gibbs sampler. Tema degli esploratori: statistica e scienza dei dati (elenco arXiv stat.ML, nuovi lavori del giorno). Categorie arXiv: Computation (stat.CO) ; Signal Processing (eess.SP); Probability (math.PR); Methodology (stat.ME); Machine Learning (stat.ML). Pagina del lavoro: https://arxiv.org/abs/2610.38753 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-01T02:13:28.607Z dalla pagina https://arxiv.org/list/stat.ML/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).