Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 24 min ago
Peer-reviewedCC BYJournal of the Royal Statistical Society Series B (Statistical Methodology)

Optimized annealed sequential Monte Carlo samplers

  • Saifuddin Syed1
  • Alexandre Bouchard‐Côté1
  • Kevin T. Chern1
  • Arnaud Doucet B2
  1. 1Department of Statistics, University of British Columbia , Vancouver, Canada
  2. 2Department of Statistics, University of Oxford, Oxford ,
Open at sourceDOICite

Abstract

Annealed sequential Monte Carlo (ASMC) samplers are special cases of SMC samplers where the sequence of distributions can be embedded in a smooth path of distributions. Using this underlying path and a performance model based on the variance of the normalizing constant estimator, we systematically study dense-schedule limits. From our theory emerges a notion of global barrier, capturing the inherent complexity of normalizing constant approximation under our performance model. We then turn the resulting approximations into surrogate objective functions of algorithm performance, using them to guide method development. This leads to novel adaptive methods, optimized annealed SMC (OASMC), which address practical difficulties inherent in previous adaptive SMC methods. First, our OASMC algorithms are predictable: they produce a sequence of increasingly precise estimates at deterministic, known times. Second, optimized annealed importance sampling (OAIS), a special case of OASMC, enables schedule adaptation at a memory cost constant in the number of particles, requiring significantly less communication. Finally, these characteristics make OAIS highly efficient on GPUs. We provide an open-source, high-performance GPU implementation of our method and demonstrate up to a hundred-fold speed improvement compared to state-of-the-art adaptive AIS methods.

Related

  1. On the convergence of dynamic implementations of Hamiltonian Monte Carlo and no U-turn samplers
    Alain Durmus, Samuel Gruffaz, Miika Kailas +2
    The Annals of Applied ProbabilityPeer-reviewedLicence not statedFull text13 citedStatisticsOpenAlex
  2. Differentiable Systematic Resampling for Variational Sequential Monte Carlo
    Fredrik Cumlin, Saikat Chatterjee
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  3. Sequince: a modular sequential Monte Carlo toolkit
    Michael J. Williams, Christian E. A. Chapman-Bird
    arXivPreprintCC BY 4.0Full textInstrumentation & methodsStatisticsarXiv
  4. Anchored Reversible Jump Sequential Monte Carlo
    Janna van Assen, Max Hinne
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  5. Parameter estimation in Conditional Sequential Monte Carlo algorithms through Particle Learning
    Alfonso Diz-Lois Palomares, Geir Storvik
    arXivPreprintCC BY-NC-ND 4.0Full textStatisticsarXiv

Cite

BibTeXRIS
BibTeX
@article{syed2026optimized,
  title = {{Optimized annealed sequential Monte Carlo samplers}},
  author = {Saifuddin Syed and Alexandre Bouchard‐Côté and Kevin T. Chern and Arnaud Doucet B},
  journal = {Journal of the Royal Statistical Society Series B (Statistical Methodology)},
  year = {2026},
  month = jun,
  doi = {10.1093/jrsssb/qkag082},
  url = {https://doi.org/10.1093/jrsssb/qkag082},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Optimized annealed sequential Monte Carlo samplers
AU  - Saifuddin Syed
AU  - Alexandre Bouchard‐Côté
AU  - Kevin T. Chern
AU  - Arnaud Doucet B
PY  - 2026
DA  - 2026/06/18
JO  - Journal of the Royal Statistical Society Series B (Statistical Methodology)
DO  - 10.1093/jrsssb/qkag082
UR  - https://doi.org/10.1093/jrsssb/qkag082
AB  - Annealed sequential Monte Carlo (ASMC) samplers are special cases of SMC samplers where the sequence of distributions can be embedded in a smooth path of distributions. Using this underlying path and a performance model based on the variance of the normalizing constant estimator, we systematically study dense-schedule limits. From our theory emerges a notion of global barrier, capturing the inherent complexity of normalizing constant approximation under our performance model. We then turn the resulting approximations into surrogate objective functions of algorithm performance, using them to guide method development. This leads to novel adaptive methods, optimized annealed SMC (OASMC), which address practical difficulties inherent in previous adaptive SMC methods. First, our OASMC algorithms are predictable: they produce a sequence of increasingly precise estimates at deterministic, known times. Second, optimized annealed importance sampling (OAIS), a special case of OASMC, enables schedule adaptation at a memory cost constant in the number of particles, requiring significantly less communication. Finally, these characteristics make OAIS highly efficient on GPUs. We provide an open-source, high-performance GPU implementation of our method and demonstrate up to a hundred-fold speed improvement compared to state-of-the-art adaptive AIS methods.
SN  - 1369-7412
SN  - 1467-9868
ER  -