pith. sign in

arxiv: 2607.01012 · v1 · pith:W5WIHWT7new · submitted 2026-07-01 · 💻 cs.LG

Generative Model Proposal based Particle Filtering for Data Assimilation

Chandni Nagda , Mayank Shrivastavam Gudrun Thorkelsdottir , Gan Zhang , Morteza Mardani , Arindam Banerjee This is my paper

Pith reviewed 2026-07-02 16:01 UTC · model grok-4.3

classification 💻 cs.LG
keywords particle filtersdata assimilationgenerative modelsproposal distributionconditional generationhigh-dimensional filteringBayesian filtering
0
0 comments X

The pith

Flow Proposal Particle Filters learn a conditional generative model to approximate the optimal proposal for particle propagation in data assimilation.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces Flow Proposal Particle Filters (FPPF) that learn a conditional generative model as the proposal distribution in particle filtering. This model conditions on observations to steer particles toward high-likelihood regions before weighting. The approach reduces weight variance and delays degeneracy while admitting tractable likelihood evaluation, which supports accurate importance weights and a full Bayesian update step. It further applies localization to handle high-dimensional problems and reports better performance than baselines on non-linear, non-Gaussian dynamical systems.

Core claim

FPPF learns a conditional generative model based proposal approximating the variance-minimizing optimal proposal for particle propagation. Conditioning on observations steers particles toward high-likelihood regions before weighting, reducing weight variance and delaying degeneracy. Since our proposal admits tractable likelihood evaluation, FPPF computes accurate importance weights and retains a Bayesian update step. We further extend FPPF to high-dimensional problems through localization strategies.

What carries the argument

The conditional generative model proposal that approximates the variance-minimizing optimal proposal while admitting tractable likelihood evaluation.

If this is right

  • Particles are steered toward high-likelihood regions before weighting, lowering variance.
  • Accurate importance weights remain available because the proposal supports tractable likelihood evaluation.
  • The Bayesian update step is retained rather than replaced by a learned approximation.
  • Localization strategies address degeneracy in high-dimensional settings.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The method may support sequential state estimation in domains such as weather modeling where observations arrive over time and state spaces are large.
  • Efficient training of the generative proposal could allow the filter to run at speeds suitable for online tracking tasks.
  • If localization introduces bias in some geometries, combining it with adaptive proposal training might mitigate that effect.

Load-bearing premise

The learned conditional generative model can be trained to sufficiently approximate the optimal proposal distribution while preserving tractable likelihood evaluation.

What would settle it

A high-dimensional non-linear dynamical system experiment where FPPF exhibits higher weight variance or faster degeneracy than a standard particle filter with the same number of particles.

Figures

Figures reproduced from arXiv: 2607.01012 by Arindam Banerjee, Chandni Nagda, Gan Zhang, Mayank Shrivastavam Gudrun Thorkelsdottir, Morteza Mardani.

Figure 1
Figure 1. Figure 1: Scaling on Lorenz-96 with arctan(x) observation operator across state dimensions dx ∈ {5, 10, 15, 20, 25, 50}. FPPF attains the lowest RMSE and CRPS at every dx, with the gap over base￾lines widening as dimension grows. Bars show means over 100 trajectories of 200 time steps and error bars denote one standard deviation. Table in Appendix C. dx. Its learned proposal directly targets high-posterior regions b… view at source ↗
Figure 2
Figure 2. Figure 2: RMSE over time for Lorenz-96 at different state dimensions. FPPF (blue) maintains lower error [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Pre-resample ESS over time for Lorenz-96 across state dimensions. Curves show rolling means [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: RMSE and CRPS on Lorenz-96 at dx ∈ {50, 100, 500, 1000} for localized and global filters under observation operator h(x) = arctan(x). L-FPPF is trained once on dx=25 and applied zero-shot at all target dimensions; localization radius r=4 is fixed across L-FPPF, LBPF, and LETKF. Bars show mean RMSE across independent runs and error bars denote one standard deviation. Table is reported in Appendix D. d=50 d=… view at source ↗
Figure 5
Figure 5. Figure 5: RMSE and CRPS on Lorenz-96 at dx ∈ {50, 100, 500, 1000} for localized and global filters under observation operator h(x) = min(x 4 , 10). L-FPPF is trained once on dx=25 and applied zero-shot at all target dimensions; localization radius r=4 is fixed across L-FPPF, LBPF, and LETKF. Bars show mean RMSE across independent runs and error bars denote one standard deviation. Table is reported in Appendix D. and… view at source ↗
Figure 6
Figure 6. Figure 6: Left: Wasserstein W2 distance from the learned proposal qϕ to the closed-form optimal proposal q ⋆ over training epochs, averaged over 500 test conditioning pairs; it decreases toward 0, while the dashed line is the bootstrap proposal’s W2 to q ⋆ . Middle: one-step ESS out of N=250 over training epochs; qϕ rises toward the maximum of 250, while the bootstrap proposal’s (dashed) is near 80. Right: bias of t… view at source ↗
read the original abstract

Data assimilation models state dynamics conditioned on sequential observations, and has wide-ranging scientific applications. In the filtering setting, the goal is to model the posterior over the current state given all observations so far. Classical solutions typically make simplifying distributional or functional assumptions, e.g., linear-Gaussian systems, which can be inaccurate in many scenarios. In principle, particle filters (PFs) remove these assumptions, yet often collapse in high dimensions. Recent generative approaches learn conditional state transitions, but without principled Bayesian updates they do not recover the correct filtering posterior and can accumulate error over long horizons. In this work, we introduce Flow Proposal Particle Filters (FPPF), which learn a conditional generative model based proposal approximating the variance-minimizing optimal proposal for particle propagation. Conditioning on observations steers particles toward high-likelihood regions before weighting, reducing weight variance and delaying degeneracy. Since our proposal admits tractable likelihood evaluation, FPPF computes accurate importance weights and retains a Bayesian update step. We further extend FPPF to high-dimensional problems through localization strategies, adressing another standard PF failure mode. Extensive experiments on a variety of dynamical systems show that FPPF outperforms statistical baselines and other generative methods in non-linear, non-Gaussian, and high-dimensional regimes.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The manuscript introduces Flow Proposal Particle Filters (FPPF), which train a conditional generative model (normalizing flow) on simulated triples (x_{t-1}, y_t, x_t) to serve as a proposal distribution approximating the optimal p(x_t | x_{t-1}, y_t) in particle filtering. Conditioning on observations steers particles toward high-likelihood regions to reduce weight variance and degeneracy; because the flow supplies an exact density, the importance weight p(y_t | x_t) p(x_t | x_{t-1}) / q remains a valid Bayesian correction. Localization is added for high-dimensional cases, and the abstract asserts that extensive experiments on varied dynamical systems demonstrate outperformance over statistical baselines and other generative methods in non-linear, non-Gaussian, and high-dimensional regimes.

Significance. If the empirical claims are substantiated, the work would offer a principled integration of modern generative models with classical sequential Monte Carlo methods, preserving theoretical correctness while addressing practical failure modes of particle filters.

major comments (2)
  1. Abstract: the assertion of outperformance on varied dynamical systems supplies no quantitative results, error bars, dataset details, or ablation studies, so the central empirical claim cannot be evaluated.
  2. Method description (throughout): no equations, derivations, or explicit training objective are shown that demonstrate how the learned conditional density q approximates the variance-minimizing optimal proposal or how the importance weights are formed, preventing verification that a Bayesian update is retained.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. The comments identify important issues with the presentation of empirical claims and methodological details. We address each major comment below and indicate the revisions we will make.

read point-by-point responses

  1. Referee: Abstract: the assertion of outperformance on varied dynamical systems supplies no quantitative results, error bars, dataset details, or ablation studies, so the central empirical claim cannot be evaluated.

    Authors: We agree that the abstract, as currently written, states the outperformance claim at a high level without supporting numbers. The Experiments section of the manuscript contains the quantitative results, error bars, dataset descriptions, and ablations, but these are not summarized in the abstract. We will revise the abstract to include a concise statement of the main quantitative improvements (e.g., relative error reductions on the tested nonlinear systems) while respecting length constraints. revision: yes

  2. Referee: Method description (throughout): no equations, derivations, or explicit training objective are shown that demonstrate how the learned conditional density q approximates the variance-minimizing optimal proposal or how the importance weights are formed, preventing verification that a Bayesian update is retained.

    Authors: The current manuscript version does not contain the requested equations or derivations. We will add a new subsection that (i) states the training objective for the conditional normalizing flow (negative log-likelihood on simulated triples (x_{t-1}, y_t, x_t) to approximate p(x_t | x_{t-1}, y_t)), (ii) derives the importance weight w_t = p(y_t | x_t) p(x_t | x_{t-1}) / q(x_t | x_{t-1}, y_t), and (iii) shows that the resulting weighted particles target the correct filtering posterior. This will make the Bayesian correction explicit and verifiable. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The derivation trains a conditional generative model via maximum likelihood on triples simulated from the known transition and observation models to target the optimal proposal p(x_t | x_{t-1}, y_t). The resulting flow q supplies an exact density, allowing standard importance weights p(y_t | x_t) p(x_t | x_{t-1}) / q to remain a valid Bayesian correction. No equations reduce performance claims to quantities defined by the fit itself, no load-bearing self-citations appear, and localization is presented as a standard extension. The construction is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; the method implicitly rests on standard particle-filter assumptions plus the existence of a trainable generative model with tractable density. No explicit free parameters, axioms or invented entities are named beyond those standard in the domain.

pith-pipeline@v0.9.1-grok · 5769 in / 1167 out tokens · 30768 ms · 2026-07-02T16:01:16.515362+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

99 extracted references · 60 canonical work pages · 7 internal anchors

  1. [1]

    Ades and P

    M. Ades and P. J. van Leeuwen. The equivalent-weights particle filter in a high-dimensional system. Quarterly Journal of the Royal Meteorological Society, 141(687):484–503, 2015. doi: 10.1002/qj.2370

    work page doi:10.1002/qj.2370 2015

  2. [2]

    X. Ai, Y . He, A. Gu, R. Salakhutdinov, J. Z. Kolter, N. M. Boffi, and M. Simchowitz. Joint distillation for fast likelihood evaluation and sampling in flow-based models. InThe Fourteenth International Con- ference on Learning Representations (ICLR), 2026. URL https://openreview.net/forum? id=8uZ5UdIul2. arXiv:2512.02636

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  3. [3]

    B. D. O. Anderson and J. B. Moore.Optimal Filtering. Prentice-Hall, Englewood Cliffs, NJ, 1979

    1979

  4. [4]

    J. L. Anderson and S. L. Anderson. A Monte Carlo implementation of the nonlinear filtering problem to produce ensemble assimilations and forecasts.Monthly Weather Review, 127(12):2741–2758, 1999. doi: 10.1175/1520-0493(1999)127<2741:AMCIOT>2.0.CO;2

    work page doi:10.1175/1520-0493(1999)127 1999

  5. [5]

    DAISI: Data Assimilation with Inverse Sampling using Stochastic Interpolants

    M. Andrae, E. Larsson, S. Takao, T. Landelius, and F. Lindsten. Daisi: Data assimilation with inverse sampling using stochastic interpolants. InProceedings of the 43rd International Conference on Machine Learning (ICML), 2026. URLhttps://arxiv.org/abs/2512.00252

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  6. [6]

    M. S. Arulampalam, S. Maskell, N. Gordon, and T. Clapp. A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking.IEEE Transactions on Signal Processing, 50(2):174–188,

  7. [7]

    doi: 10.1109/78.978374

    work page doi:10.1109/78.978374

  8. [8]

    F. Bao, Z. Zhang, and G. Zhang. An ensemble score filter for tracking high-dimensional nonlinear dynamical systems.Computer Methods in Applied Mechanics and Engineering, 432:117447, Dec. 2024. ISSN 0045-7825. doi: 10.1016/j.cma.2024.117447

    work page doi:10.1016/j.cma.2024.117447 2024

  9. [9]

    Wiley-Interscience

    Y . Bar-Shalom, X. R. Li, and T. Kirubarajan.Estimation with Applications to Tracking and Navigation: Theory, Algorithms and Software. John Wiley & Sons, New York, 2001. ISBN 9780471416555. doi: 10.1002/0471221279

    work page doi:10.1002/0471221279 2001

  10. [10]

    Bengtsson, P

    T. Bengtsson, P. Bickel, and B. Li.Curse-of-dimensionality revisited: Collapse of the particle fil- ter in very large scale systems, page 316–334. Institute of Mathematical Statistics, 2008. ISBN 19 0940600749. doi: 10.1214/193940307000000518. URL http://dx.doi.org/10.1214/ 193940307000000518

    work page doi:10.1214/193940307000000518 2008

  11. [11]

    C. H. Bishop, B. J. Etherton, and S. J. Majumdar. Adaptive sampling with the ensemble transform Kalman filter. Part I: Theoretical aspects.Monthly Weather Review, 129(3):420–436, 2001. doi: 10.1175/1520-0493(2001)129<0420:ASWTET>2.0.CO;2

    work page doi:10.1175/1520-0493(2001)129 2001

  12. [12]

    Bocquet, J

    M. Bocquet, J. Brajard, A. Carrassi, and L. Bertino. Bayesian inference of chaotic dynamics by merging data assimilation, machine learning and expectation-maximization.Foundations of Data Science, 2(1): 55–80, 2020. doi: 10.3934/fods.2020004

    work page doi:10.3934/fods.2020004 2020

  13. [13]

    Brajard, A

    J. Brajard, A. Carrassi, M. Bocquet, and L. Bertino. Combining data assimilation and machine learning to emulate a dynamical model from sparse and noisy observations: A case study with the Lorenz 96 model.Journal of Computational Science, 44:101171, 2020. doi: 10.1016/j.jocs.2020.101171

    work page doi:10.1016/j.jocs.2020.101171 2020

  14. [14]

    Branchini and V

    N. Branchini and V . Elvira. Optimized auxiliary particle filters: adapting mixture proposals via convex optimization. InConference on Uncertainty in Artificial Intelligence, 2020. URL https: //api.semanticscholar.org/CorpusID:235446306

    2020

  15. [15]

    S. Chen, Y . Jia, Q. Qu, H. Sun, and J. A. Fessler. FlowDAS: A stochastic interpolant-based framework for data assimilation. InThe Thirty-ninth Annual Conference on Neural Information Processing Systems,

  16. [16]

    URLhttps://openreview.net/forum?id=1nWqhiulqD

  17. [17]

    Chen and Y

    X. Chen and Y . Li. An overview of differentiable particle filters for data-adaptive sequential Bayesian inference.Foundations of Data Science, 2023. doi: 10.3934/fods.2023014

    work page doi:10.3934/fods.2023014 2023

  18. [18]

    Chen and Y

    X. Chen and Y . Li. Normalizing flow-based differentiable particle filters.IEEE Transactions on Signal Processing, 2024. arXiv:2403.01499

    work page arXiv 2024

  19. [19]

    Y . Chen, D. Sanz-Alonso, and R. Willett. Autodifferentiable ensemble Kalman filters.SIAM Journal on Mathematics of Data Science, 4(2):801–833, 2022. doi: 10.1137/21M1434477

    work page doi:10.1137/21m1434477 2022

  20. [20]

    Z. Chen. Bayesian filtering: From kalman filters to particle filters, and beyond.Statistics, 182(1):1–69,

  21. [21]

    doi: 10.1080/02331880309257

    work page doi:10.1080/02331880309257

  22. [22]

    A. J. Chorin and X. Tu. Implicit sampling for particle filters.Proceedings of the National Academy of Sciences, 106(41):17249–17254, 2009. doi: 10.1073/pnas.0909196106

    work page doi:10.1073/pnas.0909196106 2009

  23. [23]

    Corenflos, J

    A. Corenflos, J. Thornton, G. Deligiannidis, and A. Doucet. Differentiable particle filtering via entropy- regularized optimal transport. InProceedings of the 38th International Conference on Machine Learning (ICML), volume 139 ofProceedings of Machine Learning Research, pages 2100–2111. PMLR, 2021

    2021

  24. [24]

    Cornebise, É

    J. Cornebise, É. Moulines, and J. Olsson. Adaptive methods for sequential importance sampling with application to state space models.Statistics and Computing, 18(4):461–480, 2008. doi: 10.1007/ s11222-008-9089-4

    2008

  25. [25]

    Courtier, J.-N

    P. Courtier, J.-N. Thépaut, and A. Hollingsworth. A strategy for operational implementation of 4D-Var, using an incremental approach.Quarterly Journal of the Royal Meteorological Society, 120(519): 1367–1387, 1994. doi: 10.1002/qj.49712051912

    work page doi:10.1002/qj.49712051912 1994

  26. [26]

    A. J. W. de Wit and C. A. van Diepen. Crop model data assimilation with the Ensemble Kalman Filter for improving regional crop yield forecasts.Agricultural and Forest Meteorology, 146(1–2):38–56,

  27. [27]

    doi: 10.1016/j.agrformet.2007.05.004. 20

    work page doi:10.1016/j.agrformet.2007.05.004 2007

  28. [28]

    D. P. Dee, S. M. Uppala, A. J. Simmons, P. Berrisford, P. Poli, S. Kobayashi, U. Andrae, M. A. Balmaseda, G. Balsamo, P. Bauer, P. Bechtold, A. C. M. Beljaars, L. van de Berg, J. Bidlot, N. Bormann, C. Delsol, R. Dragani, M. Fuentes, A. J. Geer, L. Haimberger, S. B. Healy, H. Hersbach, E. V . Hólm, L. Isaksen, P. Kållberg, M. Köhler, M. Matricardi, A. P. ...

    work page doi:10.1002/qj.828 2011

  29. [29]

    Dellaert, D

    F. Dellaert, D. Fox, W. Burgard, and S. Thrun. Monte carlo localization for mobile robots. In Proceedings 1999 IEEE International Conference on Robotics and Automation (Cat. No.99CH36288C), volume 2, pages 1322–1328 vol.2, 1999. doi: 10.1109/ROBOT.1999.772544

    work page doi:10.1109/robot.1999.772544 1999

  30. [30]

    Dlamini, O

    L. Dlamini, O. Crespo, J. van Dam, and L. Kooistra. A global systematic review of improving crop model estimations by assimilating remote sensing data: Implications for small-scale agricultural systems.Remote Sensing, 15(16), 2023. ISSN 2072-4292. doi: 10.3390/rs15164066. URL https: //www.mdpi.com/2072-4292/15/16/4066

    work page doi:10.3390/rs15164066 2023

  31. [31]

    Doucet and A

    A. Doucet and A. M. Johansen. A tutorial on particle filtering and smoothing: Fifteen years later. In D. Crisan and B. Rozovskii, editors,The Oxford Handbook of Nonlinear Filtering, pages 656–704. Oxford University Press, Oxford, 2011

    2011

  32. [32]

    Doucet, S

    A. Doucet, S. Godsill, and C. Andrieu. On sequential Monte Carlo sampling methods for Bayesian filtering.Statistics and Computing, 10(3):197–208, 2000

    2000

  33. [33]

    Doucet, N

    A. Doucet, N. de Freitas, and N. Gordon.Sequential Monte Carlo Methods in Practice. Springer, New York, 2001. doi: 10.1007/978-1-4757-3437-9

    work page doi:10.1007/978-1-4757-3437-9 2001

  34. [34]

    Drton and M

    M. Drton and M. H. Maathuis. Structure learning in graphical modeling.Annual Review of Statistics and Its Application, 4(1):365–393, 2017. doi: 10.1146/annurev-statistics-060116-053803

    work page doi:10.1146/annurev-statistics-060116-053803 2017

  35. [35]

    G. Evensen. Sequential data assimilation with a nonlinear quasi-geostrophic model using monte carlo methods.Journal of Geophysical Research: Oceans, 99(C5):10143–10162, 1994. doi: 10.1029/ 94JC00572. URLhttps://doi.org/10.1029/94JC00572

    work page doi:10.1029/94jc00572 1994

  36. [36]

    Farchi and M

    A. Farchi and M. Bocquet. Review article: Comparison of local particle filters and new implementations. Nonlinear Processes in Geophysics, 25(4):765–807, 2018. doi: 10.5194/npg-25-765-2018

    work page doi:10.5194/npg-25-765-2018 2018

  37. [37]

    One Step Diffusion via Shortcut Models

    K. Frans, D. Hafner, S. Levine, and P. Abbeel. One step diffusion via shortcut models. InInternational Conference on Learning Representations (ICLR), 2025. arXiv:2410.12557

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  38. [38]

    Gaspari and S

    G. Gaspari and S. E. Cohn. Construction of correlation functions in two and three dimensions.Quarterly Journal of the Royal Meteorological Society, 125(554):723–757, 1999. doi: 10.1002/qj.49712555417

    work page doi:10.1002/qj.49712555417 1999

  39. [39]

    Z. Geng, M. Deng, X. Bai, J. Z. Kolter, and K. He. Mean flows for one-step generative modeling. In Advances in Neural Information Processing Systems, volume 38, 2025. arXiv:2505.13447

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  40. [40]

    Gordon, D

    N. Gordon, D. Salmond, and A. Smith. Novel approach to nonlinear/non-gaussian bayesian state estimation.IEE Proceedings F (Radar and Signal Processing), 140:107–113, 1993. doi: 10.1049/ ip-f-2.1993.0015. URL https://digital-library.theiet.org/doi/abs/10.1049/ ip-f-2.1993.0015

    work page arXiv 1993

  41. [41]

    G. A. Gottwald, S. Liu, Y . Marzouk, S. Reich, and X. T. Tong. Localized diffusion models.arXiv preprint arXiv:2505.04417, 2025. doi: 10.48550/arXiv.2505.04417. 21

    work page doi:10.48550/arxiv.2505.04417 2025

  42. [42]

    S. S. Gu, Z. Ghahramani, and R. E. Turner. Neural adaptive sequential Monte Carlo. InAdvances in Neural Information Processing Systems, volume 28, 2015

    2015

  43. [43]

    Hersbach, B

    H. Hersbach, B. Bell, P. Berrisford, S. Hirahara, A. Horányi, J. Muñoz-Sabater, J. Nicolas, C. Peubey, R. Radu, D. Schepers, A. Simmons, C. Soci, S. Abdalla, X. Abellan, G. Balsamo, P. Bechtold, G. Biavati, J. Bidlot, M. Bonavita, G. De Chiara, P. Dahlgren, D. Dee, M. Diamantakis, R. Dragani, J. Flemming, R. Forbes, M. Fuentes, A. Geer, L. Haimberger, S. ...

    work page doi:10.1002/qj.3803 1999

  44. [44]

    Ho and T

    J. Ho and T. Salimans. Classifier-free diffusion guidance. InNeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications, 2021. URL https://openreview.net/ forum?id=qw8AKxfYbI

    2021

  45. [45]

    P. L. Houtekamer and H. L. Mitchell. A sequential ensemble kalman filter for atmospheric data assimilation.Monthly Weather Review, 129(1):123 – 137, 2001. doi: 10.1175/1520-0493(2001) 129<0123:ASEKFF>2.0.CO;2. URL https://journals.ametsoc.org/view/journals/ mwre/129/1/1520-0493_2001_129_0123_asekff_2.0.co_2.xml

    work page doi:10.1175/1520-0493(2001 2001

  46. [46]

    P. L. Houtekamer, X. Deng, H. L. Mitchell, S.-J. Baek, and N. Gagnon. Higher resolution in an operational ensemble kalman filter.Monthly Weather Review, 142(3):1143 – 1162, 2014. doi: 10.1175/ MWR-D-13-00138.1. URL https://journals.ametsoc.org/view/journals/mwre/ 142/3/mwr-d-13-00138.1.xml

    2014

  47. [47]

    P. L. Houtekamer, X. Deng, H. L. Mitchell, S.-J. Baek, and N. Gagnon. Higher resolution in an operational ensemble Kalman filter.Monthly Weather Review, 142:1143–1162, 2014

    2014

  48. [48]

    Huang, L

    L. Huang, L. Gianinazzi, Y . Yu, P. D. Dueben, and T. Hoefler. DiffDA: a diffusion model for weather- scale data assimilation. InForty-first International Conference on Machine Learning, 2024. URL https://openreview.net/forum?id=vhMq3eAB34

    2024

  49. [49]

    B. R. Hunt, E. J. Kostelich, and I. Szunyogh. Efficient data assimilation for spatiotemporal chaos: A local ensemble transform Kalman filter.Physica D: Nonlinear Phenomena, 230(1–2):112–126, 2007. doi: 10.1016/j.physd.2006.11.008

    work page doi:10.1016/j.physd.2006.11.008 2007

  50. [50]

    M. F. Hutchinson. A stochastic estimator of the trace of the influence matrix for Laplacian smoothing splines.Communications in Statistics - Simulation and Computation, 19(2):433–450, 1990. doi: 10.1080/03610919008812866

    work page doi:10.1080/03610919008812866 1990

  51. [51]

    A. H. Jazwinski.Stochastic Processes and Filtering Theory, volume 64 ofMathematics in Science and Engineering. Academic Press, New York, 1970

    1970

  52. [52]

    Jonschkowski, D

    R. Jonschkowski, D. Rastogi, and O. Brock. Differentiable particle filters: End-to-end learning with algorithmic priors. InProceedings of Robotics: Science and Systems (RSS), Pittsburgh, Pennsylvania,

  53. [53]

    doi: 10.15607/RSS.2018.XIV .001

    work page doi:10.15607/rss.2018.xiv 2018

  54. [54]

    S. J. Julier and J. K. Uhlmann. New extension of the Kalman filter to nonlinear systems. InSignal Processing, Sensor Fusion, and Target Recognition VI, volume 3068 ofProc. SPIE, pages 182–193,

  55. [55]

    doi: 10.1117/12.280797

    work page doi:10.1117/12.280797

  56. [56]

    R. E. Kalman. A new approach to linear filtering and prediction problems.Journal of Basic Engineering, 82(1):35–45, 1960. doi: 10.1115/1.3662552. 22

    work page doi:10.1115/1.3662552 1960

  57. [57]

    Karkus, D

    P. Karkus, D. Hsu, and W. S. Lee. Particle filter networks with application to visual localization. In Proceedings of the 2nd Conference on Robot Learning (CoRL), volume 87 ofProceedings of Machine Learning Research, pages 169–178. PMLR, 2018

    2018

  58. [58]

    Kuramoto

    Y . Kuramoto. Diffusion-induced chaos in reaction systems.Progress of Theoretical Physics Supplement, 64:346–367, 1978. doi: 10.1143/PTPS.64.346

    work page doi:10.1143/ptps.64.346 1978

  59. [59]

    S. L. Lauritzen.Graphical Models. Number 17 in Oxford Statistical Science Series. Clarendon Press, Oxford, 1996

    1996

  60. [60]

    T. A. Le, M. Igl, T. Rainforth, T. Jin, and F. Wood. Auto-encoding sequential Monte Carlo. In International Conference on Learning Representations, 2018. URL https://openreview.net/ forum?id=BJ8c3f-0b

    2018

  61. [61]

    Lipman, R

    Y . Lipman, R. T. Q. Chen, H. Ben-Hamu, M. Nickel, and M. Le. Flow matching for generative modeling. InThe Eleventh International Conference on Learning Representations, 2023. URL https://openreview.net/forum?id=PqvMRDCJT9t

    2023

  62. [62]

    Flow Matching Guide and Code

    Y . Lipman, M. Havasi, P. Holderrieth, N. Shaul, M. Le, B. Karrer, R. T. Q. Chen, D. Lopez-Paz, H. Ben-Hamu, and I. Gat. Flow matching guide and code, 2024. URL https://arxiv.org/abs/ 2412.06264

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  63. [63]

    J. S. Liu and R. Chen. Sequential Monte Carlo methods for dynamic systems.Journal of the American Statistical Association, 93(443):1032–1044, 1998. doi: 10.1080/01621459.1998.10473765

    work page doi:10.1080/01621459.1998.10473765 1998

  64. [64]

    X. Liu, C. Gong, and qiang liu. Flow straight and fast: Learning to generate and transfer data with rectified flow. InThe Eleventh International Conference on Learning Representations, 2023. URL https://openreview.net/forum?id=XVjTT1nw5z

    2023

  65. [65]

    A. C. Lorenc. Analysis methods for numerical weather prediction.Quarterly Journal of the Royal Meteorological Society, 112(474):1177–1194, 1986. doi: 10.1002/qj.49711247414

    work page doi:10.1002/qj.49711247414 1986

  66. [66]

    A. C. Lorenc. Analysis methods for numerical weather prediction.Quarterly Journal of the Royal Meteorological Society, 112(474):1177–1194, 1986

    1986

  67. [69]

    E. N. Lorenz. Deterministic nonperiodic flow.Journal of the Atmospheric Sciences, 20(2):130–141,

  68. [70]

    doi: 10.1175/1520-0469(1963)020<0130:DNF>2.0.CO;2

    work page doi:10.1175/1520-0469(1963)020 1963

  69. [71]

    E. N. Lorenz. Predictability: A problem partly solved. InProceedings of Seminar on Predictability, pages 1–18, Reading, UK, 1995. ECMWF

    1995

  70. [72]

    E. N. Lorenz. Predictability: A problem partly solved. InProc. Seminar on Predictability, volume 1, pages 1–18, Reading, UK, 1996. ECMWF

    1996

  71. [73]

    C. J. Maddison, D. Lawson, G. Tucker, N. Heess, M. Norouzi, A. Mnih, A. Doucet, and Y . W. Teh. Filtering variational objectives. InAdvances in Neural Information Processing Systems, volume 30, 2017

    2017

  72. [74]

    J. E. Matheson and R. L. Winkler. Scoring rules for continuous probability distributions.Management Science, 22(10):1087–1096, 1976. doi: 10.1287/mnsc.22.10.1087. 23

    work page doi:10.1287/mnsc.22.10.1087 1976

  73. [75]

    Morzfeld, X

    M. Morzfeld, X. Tu, E. Atkins, and A. J. Chorin. A random map implementation of implicit filters. Journal of Computational Physics, 231(4):2049–2066, 2012. doi: 10.1016/j.jcp.2011.11.022

    work page doi:10.1016/j.jcp.2011.11.022 2049

  74. [76]

    C. A. Naesseth, S. W. Linderman, R. Ranganath, and D. M. Blei. Variational sequential Monte Carlo. InProceedings of the 21st International Conference on Artificial Intelligence and Statistics (AISTATS), volume 84 ofProceedings of Machine Learning Research, pages 968–977. PMLR, 2018

    2018

  75. [77]

    Paige and F

    B. Paige and F. Wood. Inference networks for sequential Monte Carlo in graphical models. In Proceedings of the 33rd International Conference on Machine Learning, volume 48 ofProceedings of Machine Learning Research, pages 3040–3049. PMLR, 2016

    2016

  76. [78]

    S. G. Penny and T. Miyoshi. A local particle filter for high-dimensional geophysical systems.Nonlinear Processes in Geophysics, 23(6):391–405, 2016. doi: 10.5194/npg-23-391-2016

    work page doi:10.5194/npg-23-391-2016 2016

  77. [79]

    Perez, F

    E. Perez, F. Strub, H. de Vries, V . Dumoulin, and A. Courville. Film: Visual reasoning with a general conditioning layer.Proceedings of the AAAI Conference on Artificial Intelligence, 32(1), Apr. 2018. doi: 10.1609/aaai.v32i1.11671. URL https://ojs.aaai.org/index.php/AAAI/ article/view/11671

    work page doi:10.1609/aaai.v32i1.11671 2018

  78. [80]

    M. K. Pitt and N. Shephard. Filtering via simulation: Auxiliary particle filters.Journal of the American Statistical Association, 94(446):590–599, 1999. doi: 10.1080/01621459.1999.10474153

    work page doi:10.1080/01621459.1999.10474153 1999

  79. [81]

    Poterjoy

    J. Poterjoy. A localized particle filter for high-dimensional nonlinear systems.Monthly Weather Review, 144(1):59–76, 2016. doi: 10.1175/MWR-D-15-0163.1

    work page doi:10.1175/mwr-d-15-0163.1 2016

  80. [82]

    Rabier, H

    F. Rabier, H. Järvinen, E. Klinker, J.-F. Mahfouf, and A. Simmons. The ECMWF operational im- plementation of four-dimensional variational assimilation. I: Experimental results with simplified physics.Quarterly Journal of the Royal Meteorological Society, 126(564):1143–1170, 2000. doi: 10.1256/smsqj.56414

    work page doi:10.1256/smsqj.56414 2000

Showing first 80 references.