Pith. sign in

REVIEW 3 major objections 4 minor 124 references

Error Analysis of Triangular Optimal Transport Maps for Filtering

T0 review · 3 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper establishes finite-sample error bounds for conditional Brenier maps and derives end-to-end posterior error rates for an optimal-transport filtering algorithm.

desk verdict First real error bounds for conditional Brenier maps and an OT-based filter; the central theorem is a genuine contribution, but the worked Legendre example likely does not deliver the advertised fast rate and the numerics use an algorithm outside the analysis. read the letter →

arxiv 2510.19283 v2 pith:LVNC4P4I submitted 2025-10-22 math.ST math.OCstat.MLstat.TH

Mohammad Al-Jarrah , Bamdad Hosseini , Niyizhen Jin , Michele Martino , Amirhossein Taghvaei This is my paper · ORCID
classification math.STmath.OCstat.MLstat.TH MSC 60G3549Q2265M3262F15
keywords conditionaloptimaltransportBreniermapsnonlinearfilteringsamplecomplexityBayesianinferencesimulation-basedempiricalprocesstheorydataassimilation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper's central project is to turn a heuristic—replacing Bayes' rule in filtering by a learned transport map—into a theorem with explicit convergence rates. It shows that a conditional Brenier map, estimated by solving an empirical dual optimal transport problem over a restricted function class, has squared-gradient error decaying like (log N / N)^(2/(2+γ)) under regularity assumptions, and at least like N^(-1/2) under weak assumptions. These are among the first quantitative statistical guarantees for conditional optimal transport maps, and they matter for simulation-based inference and Bayesian conditioning. The paper then feeds these rates into a filtering recursion, bounding the posterior error of an optimal-transport filter with a slow N^(-1/4) rate and a fast (log N / N)^(1/2) rate, provided the true filter is uniformly geometrically stable. If correct, practitioners get concrete sample-size guidance and a principled alternative to bootstrap particle filters.

What carries the argument

The central object is the conditional Brenier map T(y,·) = ∇_u φ†(y,·), which pushes the reference measure η_U to the conditional posterior ν(·|y). The potential φ† is the minimizer of the conditional Kantorovich dual S(φ) = ∫[φ(y,v) + φ*(y,u)] dη dν. The analysis hinges on a stability lemma: the excess dual risk S(φ)−S(φ†) is sandwiched between (2β_max)^{-1} and (2α_min)^{-1} times the squared L² gradient error, so statistical control of the empirical dual objective implies control of the transport map itself. Empirical process chaining and a one-shot localization argument produce the N rates. In the filtering half, the same map is recomputed at every time step from particles, and uniform g

What would settle it

Run the OT filter on a hidden Markov model where the filter is not uniformly stable—for example, nearly uninformative observations with a transition kernel that lacks minorization—and test whether the empirical posterior error fails to decay with N as predicted. More directly, in a static conditional OT problem with log-concave conditionals, compute the empirical squared gradient error for N = 10^3, 10^4, 10^5 and check whether it follows (log N / N)^(2/(2+γ)); if it stalls at N^(-1/2) under conditions designed to trigger the fast rate, the localization argument or Poincaré assumptions are ins

Watch

Extended reading notes

Core claim

The load-bearing result is Theorem 2.10: under strong convexity and smoothness of the model class, Poincaré inequalities on the reference and target conditionals, and a class rich enough to approximate the true potential to order (log N / N)^(1/(2+γ)), the empirical conditional Brenier potential achieves E‖∇φ̂ − ∇φ†‖²_{L²η} ≤ C (log N / N)^(2/(2+γ)). Reading the gradient as the transport map, this is a finite-sample, parameterized estimator converging at nearly the parametric rate. The paper also proves a slow rate, Theorem 2.7, requiring only boundedness and covering-number control of the function class. For filtering, Theorems 3.4 and 3.7 convert these map errors into posterior-divergence

Load-bearing premise

The load-bearing premise is that the true filtering map forgets its starting distribution at a uniform geometric rate no matter what observations arrive—a condition the paper itself calls relatively strong and hard to verify outside compact state spaces; for the exact-error theorem, an additional fragile premise is that the conditional Brenier maps are uniformly Lipschitz in the observation, which is currently open for noncompact observation spaces.

Editorial extensions

If this is right

  • Conditional OT maps used in simulation-based inference inherit a sample complexity: N^(-1/2) squared-gradient error under weak assumptions, and (log N / N)^(2/(2+γ)) under log-concavity and Poincaré conditions.
  • OT-based filters get end-to-end guarantees: expected posterior divergence decays at N^(-1/4) generally and (log N / N)^(1/2) under the fast-rate conditions, with errors accumulating only through a geometrically discounted sum.
  • The theory identifies the quantities practitioners should monitor: the filter stability constant λ, strong convexity and smoothness bounds of the potential class, Poincaré constants, and Lipschitz constants in the observation coordinate.
  • For systems with uniformly log-concave updates and quadratic-in-u potentials, the required regularity conditions persist over time, so the fast rates can be maintained uniformly in time.
  • An EnKF-based reference measure makes the learned map near-identity in near-Gaussian settings; numerical benchmarks show the OT filter matching or beating common filters on Lorenz 63 and Lorenz 96.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The uniform geometric stability assumption is likely the main practical barrier; the paper notes it holds only for restricted systems, such as compact state spaces. A natural extension the authors leave open is replacing it with expectation-style or detectability-based stability to reach broader filtering models.
  • The exact-error rate is bottlenecked by Lipschitzness of Brenier maps in the observation y, which is open on noncompact domains. A testable extension is to prove or disprove a local or Hölder version; the paper shows Hölder exponents below 1 make its Grönwall argument diverge.
  • Because the fast rate is driven by localization, one could check whether (log N / N)^(2/(2+γ)) is minimax for conditional OT map estimation by adapting known lower bounds for unconditional OT maps; the paper does not address optimality.
  • The EnKF-reference trick suggests a broader design principle: choose a data-dependent reference close to the target so the learned correction is small, potentially extending to other linear-Gaussian approximations and reducing training difficulty.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper develops finite-sample error bounds for empirical conditional Brenier maps and applies them to an optimal-transport-based filter. Its main theoretical contributions are Theorem 2.7 (a slow O(N^{-1/2}) bound on the squared L^2 gradient error of the estimated conditional potential) and Theorem 2.10 (a fast O((log N/N)^{2/(2+γ)}) bound under additional Poincaré, smoothness, and sieve-bias assumptions). These map-estimation results are then used in Section 3 to bound the error of an idealized optimal-transport filter: Theorem 3.4 controls an approximate mean filtering error, and Theorem 3.7 extends the bound to the exact mean filtering error under filter stability and Lipschitz regularity conditions. Numerical experiments on bimodal, Lorenz-63, and Lorenz-96 benchmarks are reported in Section 4. The supplementary materials contain the detailed proofs of Theorems 2.7, 2.10, 3.4, and 3.7.

Significance. If the results hold as stated, the paper would constitute a valuable first quantitative analysis of conditional OT map estimation and of OT-based filters, extending the program of Hütter–Rigollet and Divol–Niles-Weed–Pooladian to the conditional/filtering setting. The proof strategy is coherent and the constants are tracked explicitly; the supplementary material is detailed, and the code is made available. However, the fast-rate result currently relies on a worked example whose entropy constant grows with the sieve size, which undermines the advertised application of Theorem 2.10. In addition, the filtering theorems apply to an idealized algorithm that is not the one implemented in the numerical section. These gaps are substantive but potentially repairable, so the manuscript is promising rather than ready.

major comments (3)
  1. [Section 2.4, Assumption 2.6(3), SM B.3] The Legendre example is not a valid verification of the fast-rate assumptions. Assumption 2.6(3) fixes C_F as a constant independent of N, but Section 2.4 states that for the Legendre class F(M) the entropy bound holds with 'γ=0 and C_F ∝ M+1' and then sets M = O((N/log N)^{1/(2+γ)}). Hence C_F grows with N. This is not just a notational issue: the localization proof in SM B.3 uses C_F as a constant. In Proposition B.8, θ(N,ε,t) contains ε N^{-1/2} sqrt(C_F) ε^{-γ/2} sqrt(log(1+ε^{-1})) and N^{-1} C_F (1-γ)^{-2}. For γ=0 and C_F ~ N^{1/2}, choosing ε ~ (log N/N)^{1/2} gives θ ~ N^{-1/4} log^{3/4} N, which is much larger than ε. The fixed-point step leading to (B.23) then cannot recover the claimed squared-error rate; it would only give at best N^{-1/4}. Thus the only worked example that is supposed to satisfy Assumption 2.9(4) does not substantiate Theorem 2.10. The proof must either tra
  2. [Section 3.4 vs. Section 4] The filtering theorems are proved for the idealized algorithm (3.26) and, with an extra resampling error, for (3.28). The text explicitly states that the interacting particle system (3.27) 'makes our theoretical analysis not applicable.' However, Algorithm 4.1 and the numerical experiments in Section 4 implement the interacting version: particles are propagated through previous optimal transport maps without regeneration, and in the paragraph around Eq. (4.4) the v_i^t are formed by shuffling the u_i^t. Consequently the Lorenz-63 and Lorenz-96 benchmarks do not test the end-to-end rates in Theorems 3.4 and 3.7. The manuscript should either state clearly in the abstract and conclusion that the error analysis covers an idealized filter and that the numerical algorithm is an interacting approximation, or provide a propagation-of-chaos or resampling-error analysis bridging the two. This is a
  3. [Theorem 3.7 and Remark 3.10] The exact mean filtering error bound (3.14) contains the factor max(1, ϱ^t) with ϱ = 2 L_Y C'_D C_D C_stab + 1 - λ. For the bound to be uniform in time, (3.16) must hold, namely 2 L_Y C'_D C_D C_stab ≤ λ. This is a genuine contractivity condition on the product of the map Lipschitz constant, the coupling constant, the stability constant of the divergence, and the filter stability rate. The numerical section does not verify this condition, and for neural-network parameterizations the constants L_Y and C'_D are not computed. While the authors acknowledge in Remark 3.10 that this may be an artifact, the theorem as stated does not provide a uniform filtering rate for the reported benchmarks. The manuscript should state this limitation prominently and, ideally, provide a nontrivial example where (3.16) is verified.
minor comments (4)
  1. [Section 1.1] The sentence 'directly apply to the time-invariant case' appears to mean 'time-varying'; please correct the typo.
  2. [Proof of Theorem 2.10, SM B.3] The text refers to 'Assumption 2.9(6)', but Assumption 2.9 has only four items. The reference should be to Assumption 2.9(4).
  3. [Section 2.4] Even in the slow-rate Theorem 2.7, the statement 'Assumption 2.6 with γ=0 and C_F ∝ M+1' conflicts with Assumption 2.6(3), which requires C_F to be fixed. For clarity, define C_F(M) and state how the bound depends on it when M is allowed to grow.
  4. [General] There are several spelling and notation issues, e.g., 'Lipshitz' for 'Lipschitz', and the notation eϕ in (B.4) is introduced as an argmin that may not exist; the text mentions a minimizing sequence but should make that explicit at first use.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Theorem 2.10 is a conditional bias-variance bound with a separately proven localization term; Assumption 2.9(4) is a standard sieve condition, not a fitted prediction, and the self-citations are not load-bearing.

full rationale

Theorem 2.10 concludes E_train ||grad_u bφ − grad_u φ†||^2_{L2η} ≤ C (log N/N)^{2/(2+γ)}. Assumption 2.9(4) assumes inf_φ∈F ||grad_u φ − grad_u φ†||_{L2η} ≤ C_bias (log N/N)^{1/(2+γ)}. Although the exponents match after squaring, the theorem does not reduce to the assumption: the empirical estimator bφ could in principle be far from the best-in-class element, and the proof supplies a separate localization variance bound (Proposition B.8 and the σ-selection argument around Eq. B.21–B.23) that is needed to close the bound. This is a bias-variance decomposition, not a definitional identity. The same holds for the filtering theorems: Lemma 3.3 is proved in SM C.1 and depends on the externally motivated filter-stability Assumption 3.1; Theorem 3.7's Grönwall step is an additional argument, not a circular citation. The paper's self-citations [2,5,61] are either the algorithms being analyzed or supporting results with external counterparts ([29]; [32,46,62]); none is the sole load-bearing justification. The Sec. 2.4 Legendre example's metric-entropy count (CF ∝ M+1 in d>1) is a correctness concern about whether the growing sieve satisfies Assumption 2.6(3), not a circularity, so I leave it for a correctness review.

Assumptions & free parameters 0 free parameters · 10 assumptions · 0 invented entities

The central theorems are conditional on a long list of regularity conditions: bounded strongly-convex smooth hypothesis classes, Poincaré inequalities, a bias-rate assumption, and uniform geometric filter stability. These are standard in empirical process theory but are strong and often hard to verify in filtering applications. No free parameters or invented entities enter; the fast rate is purchased with Assumption 2.9(4).

assumptions (10)
  • domain assumption Assumption 2.4: ηU has a density with convex support, conditional targets ν(·|y) admit densities, and η,ν have finite second U-moments.
    Invoked via Proposition 2.5 to ensure the conditional Brenier potential exists and is the unique minimizer of the dual problem (2.5).
  • domain assumption Assumption 2.6: F is uniformly L∞-bounded, every φ∈F is α-strongly convex and β-smooth in u, and the log-covering number of F grows like δ^{-γ}.
    Needed for the stability lemma B.1 and the Dudley/chaining bounds in Theorem 2.7; restrictive for neural network classes.
  • domain assumption Assumption 2.9(1): ηU, νY, and ν(·|y) satisfy Poincaré inequalities.
    Used in Proposition B.10 to convert L2 bounds on gradients into L2 bounds on potentials and conjugates, required for the fast rate.
  • domain assumption Assumption 2.9(2): uniform bound on mixed derivatives ∇uyφ for φ∈F∪{φ†}.
    Used to control the y-derivative of F(y) in (B.27) and in the Lipschitz transfer of Theorem 3.7; the authors note this is open for non-compact domains.
  • domain assumption Assumption 2.9(3): F is convex.
    Used in the one-shot localization argument to form the convex combination φσ; the authors note it can be removed with a more technical proof.
  • ad hoc to paper Assumption 2.9(4): inf_F ||∇φ−∇φ†|| ≤ C_bias (log N / N)^{1/(2+γ)} for N>N0.
    This presumes the map class already approximates the truth at the fast rate that the theorem then proves for the estimator. It is the most consequential assumption in the fast-rate result.
  • domain assumption Assumption 3.1: the true filter is uniformly geometrically stable, D(T_{Yt,τ}[π1], T_{Yt,τ}[π2]) ≤ C_stab(1−λ)^{t−τ+1}D(π1,π2) uniformly in observations.
    Needed for Lemma 3.3 to prevent error accumulation over time. The authors state it is relatively strong and verified only for restricted systems such as compact state spaces.
  • domain assumption Assumption 3.2: D is uniformly stable over P2(U) and satisfies the triangle inequality.
    Allows transport-map error to be transferred to divergence between pushed-forward measures; verified in SM A for W2 and MMD.
  • ad hoc to paper Assumption 3.6: ∇uφ†τ and ∇u bφτ are LY-Lipschitz in y, and the coupling bound E||Y1−Y2|| ≤ C'_D D(μ1, μ2) holds.
    Needed to convert the approximate filtering error (3.10) into the exact error (3.9). The resulting bound is time-uniform only if 2LY C'_D C_D C_stab ≤ λ, which the authors concede is strict.
  • standard math External regularity results: Caffarelli contraction, [57, Cor. 1.2] for linearized Monge-Ampère regularity, Brascamp-Lieb and Cramér-Rao inequalities, and empirical-process chaining/Dudley bounds.
    Invoked without proof in Sections 2.4, 3.3, and the supplementary materials; standard in the literature but not derived in this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Error Analysis of Triangular Optimal Transport Maps for Filtering." pith.science (2026). https://pith.science/paper/LVNC4P4I

@misc{pith2026251019283,
  author       = {Pith},
  title        = {Pith review of: Error Analysis of Triangular Optimal Transport Maps for Filtering},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LVNC4P4I}},
  note         = {Machine review of arXiv:2510.19283}
}
read the original abstract

We present a systematic analysis of estimation errors for a class of optimal transport based algorithms for filtering and data assimilation. Along the way, we extend previous error analyses of Brenier maps to the case of conditional Brenier maps that arise in the context of simulation based inference. We then apply these results in a filtering scenario to analyze the optimal transport filtering algorithm of Al-Jarrah et al. (2024, ICML). An extension of that algorithm along with numerical benchmarks on various non-Gaussian and high-dimensional examples are provided to demonstrate its effectiveness and practical potential.

Figures

Figures reproduced from arXiv: 2510.19283 by the authors.

Figure 1
Figure 1. The left figure shows the kernel density estimate function of the transported particles [PITH_FULL_IMAGE:figures/full_fig_p025_1.png] view at source ↗
Figure 2
Figure 2. The left figure shows the W2 distance as a function of dimension n for a fixed number of particles N = 5000. The middle figure shows the corresponding computational time as a function of dimension. The right figure shows the W2 distance as a function of the number of particles N for a fixed dimension n = 10. 4.3. Benchmark Lorenz 63 model. In this numerical experiment, we illustrate the per￾formance of the same nonl… view at source ↗
Figure 3
Figure 3. Numerical results for the Lorenz 63 example. The left three panels illustrate the true [PITH_FULL_IMAGE:figures/full_fig_p027_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Numerical results for the Lorenz 96 example. The left three panels depict the [PITH_FULL_IMAGE:figures/full_fig_p027_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

124 extracted references · 2 linked inside Pith

  1. [1]

    Adcock, S

    B. Adcock, S. Brugiapaglia, and C. G. Webster , Sparse polynomial approximation of high- dimensional functions , vol. 25, SIAM, 2022

    2022

  2. [2]

    Al-Jarrah, B

    M. Al-Jarrah, B. Hosseini, and A. Taghvaei, Optimal transport particle filters, in 2023 62nd IEEE Conference on Decision and Control (CDC), IEEE, 2023, pp. 6798–6805

    2023

  3. [3]

    Al-Jarrah, B

    M. Al-Jarrah, B. Hosseini, and A. Taghvaei , Data-driven approximation of stationary nonlinear filters with optimal transport maps , in 2024 IEEE 63rd Conference on Decision and Control (CDC), IEEE, 2024, pp. 2727–2733

    2024

  4. [4]

    Al-Jarrah, B

    M. Al-Jarrah, B. Hosseini, and A. Taghvaei, Fast filtering of non-Gaussian models using amortized optimal transport maps , IEEE Control Systems Letters, 9 (2025), pp. 1670–1675

    2025

  5. [5]

    Al-Jarrah, N

    M. Al-Jarrah, N. Jin, B. Hosseini, and A. Taghvaei , Nonlinear filtering with Brenier optimal transport maps, in Forty-first International Conference on Machine Learning, 2024

    2024

  6. [6]

    B. Amos, L. Xu, and J. Z. Kolter , Input convex neural networks , in Proceedings of the 34th Inter- national Conference on Machine Learning, D. Precup and Y. W. Teh, eds., vol. 70 of Proceedings of Machine Learning Research, PMLR, 06–11 Aug 2017, pp. 146–155

    2017

  7. [7]

    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 (2002), pp. 174–188

    2002

  8. [8]

    Baptista, B

    R. Baptista, B. Bloem-Reddy, and B. Hosseini , Transport for conditional sampling: Foundations and applications

Show all 124 references
  1. [9]

    Baptista, B

    R. Baptista, B. Hosseini, N. B. Kovachki, Y. Marzouk, and A. Sagiv , An approximation theory framework for measure-transport sampling algorithms , Mathematics of Computation, 94 (2025), pp. 1863–1909

    2025

  2. [10]

    Baptista, B

    R. Baptista, B. Hosseini, N. B. Kovachki, and Y. M. Marzouk, Conditional sampling with mono- tone GANs: From generative models to likelihood-free inference, SIAM/ASA Journal on Uncertainty ERROR ANALYSIS OF TRIANGULAR OPTIMAL TRANSPORT MAPS FOR FILTERING 29 Quantification, 12 ...

    2024

  3. [11]

    Baptista, A.-A

    R. Baptista, A.-A. Pooladian, M. Brennan, Y. Marzouk, and J. Niles-Weed , Conditional simulation via entropic optimal transport: Toward non-parametric estimation of conditional Brenier maps, in The 28th International Conference on Artificial Intelligence and Statistics, 2025

    2025

  4. [12]

    Bar-Shalom, X

    Y. Bar-Shalom, X. R. Li, and T. Kirubarajan , Estimation with applications to tracking and navi- gation: theory algorithms and software , John Wiley & Sons, 2004

    2004

  5. [13]

    Bengtsson, P

    T. Bengtsson, P. Bickel, and B. Li , Curse of dimensionality revisited: Collapse of the particle filter in very large scale systems , in IMS Lecture Notes - Monograph Series in Probability and Statistics: Essays in Honor of David F. Freedman, vol. 2, Institute of Mathematical ...

    2008

  6. [14]

    Beskos, D

    A. Beskos, D. Crisan, A. Jasra, and N. Whiteley , Error bounds and normalising constants for sequential Monte Carlo samplers in high dimensions , Advances in Applied Probability, 46 (2014), pp. 279–306

    2014

  7. [15]

    Bickel, B

    P. Bickel, B. Li, T. Bengtsson, et al. , Sharp failure rates for the bootstrap particle filter in high dimensions, in Pushing the limits of contemporary statistics: Contributions in honor of Jayanta K. Ghosh, Institute of Mathematical Statistics, 2008, pp. 318–329

    2008

  8. [16]

    A. N. Bishop and P. Del Moral , On the stability of matrix-valued Riccati diffusions , Electronic Journal of Probability, 24 (2019)

    2019

  9. [17]

    C. M. Bishop and N. M. Nasrabadi, Pattern recognition and machine learning, vol. 4, Springer, 2006

    2006

  10. [18]

    V. I. Bogachev and M. A. S. Ruas , Measure theory, vol. 2, Springer, 2007

    2007

  11. [19]

    S. P. Boyd and L. V andenberghe, Convex optimization , Cambridge university press, 2004

    2004

  12. [20]

    H. J. Brascamp and E. H. Lieb , On extensions of the Brunn-Minkowski and Pr´ ekopa-Leindler the- orems, including inequalities for log concave functions, and with an application to the diffusion equation, Journal of functional analysis, 22 (1976), pp. 366–389

    1976

  13. [21]

    Brigo and B

    D. Brigo and B. Hanzon , On some filtering problems arising in mathematical finance , Insurance: Mathematics and Economics, 22 (1998), pp. 53–64

    1998

  14. [22]

    Budhiraja, L

    A. Budhiraja, L. Chen, and C. Lee , A survey of numerical methods for nonlinear filtering problems , Physica D: Nonlinear Phenomena, 230 (2007), pp. 27–36

    2007

  15. [23]

    Bunne, A

    C. Bunne, A. Krause, and M. Cuturi , Supervised training of conditional Monge maps , Advances in Neural Information Processing Systems, 35 (2022), pp. 6859–6872

    2022

  16. [24]

    L. A. Caffarelli, The regularity of mappings with a convex potential , Journal of the American Math- ematical Society, 5 (1992), pp. 99–104

    1992

  17. [25]

    L. A. Caffarelli, Monotonicity properties of optimal transportation and the FKG and related inequal- ities, Communications in Mathematical Physics, 214 (2000), pp. 547–563

    2000

  18. [26]

    Calvello, S

    E. Calvello, S. Reich, and A. M. Stuart , Ensemble Kalman methods: a mean-field perspective , Acta Numerica, 34 (2025), pp. 123–291

    2025

  19. [27]

    Canuto and A

    C. Canuto and A. Quarteroni , Approximation results for orthogonal polynomials in Sobolev spaces , Mathematics of Computation, 38 (1982), pp. 67–86

    1982

  20. [28]

    Capp ´e, E

    O. Capp ´e, E. Moulines, and T. Ryd ´en, Inference in hidden Markov models , in Proceedings of EUSFLAT Conference, 2009, pp. 14–16

    2009

  21. [29]

    Carlier, V

    G. Carlier, V. Chernozhukov, and A. Galichon , Vector quantile regression: an optimal transport approach, The Annals of Statistics, 44 (2016), pp. 1165–1192

    2016

  22. [30]

    Chemseddine, P

    J. Chemseddine, P. Hagemann, G. Steidl, and C. W ald , Conditional Wasserstein distances with applications in Bayesian ot flow matching , Journal of Machine Learning Research, 26 (2025), pp. 1– 47

    2025

  23. [31]

    Chen et al

    Z. Chen et al. , Bayesian filtering: From Kalman filters to particle filters, and beyond , Statistics, 182 (2003), pp. 1–69

    2003

  24. [32]

    Chewi, J

    S. Chewi, J. Niles-Weed, and P. Rigollet , Statistical optimal transport, Springer, 2024

    2024

  25. [33]

    Chigansky, R

    P. Chigansky, R. Liptser, and R. V an Handel , Intrinsic methods in filter stability , Handbook of Nonlinear Filtering, (2009)

    2009

  26. [34]

    T. A. Courtade, Bounds on the Poincar´ e constant for convolution measures, (2020)

    2020

  27. [35]

    Cranmer, J

    K. Cranmer, J. Brehmer, and G. Louppe , The frontier of simulation-based inference , Proceedings of the National Academy of Sciences, 117 (2020), pp. 30055–30062

    2020

  28. [36]

    Crisan and B

    D. Crisan and B. Rozovskii , The Oxford handbook of nonlinear filtering , Oxford University Press, 2011. 30 M. AL-JARRAH, B. HOSSEINI, N. JIN, M. MARTINO, A. TAGHVAEI

    2011

  29. [37]

    Crisan and J

    D. Crisan and J. Xiong, Approximate McKean-Vlasov representations for a class of SPDEs , Stochas- tics An International Journal of Probability and Stochastics Processes, 82 (2010), pp. 53–68

    2010

  30. [38]

    Dasgupta, H

    A. Dasgupta, H. Ramaswamy, J. Murgoitio-Esandi, K. Y. Foo, R. Li, Q. Zhou, B. F. Kennedy, and A. A. Oberai, Conditional score-based diffusion models for solving inverse elasticity problems , Computer Methods in Applied Mechanics and Engineering, 433 (2025), p. 117425

    2025

  31. [39]

    F. Daum, J. Huang, and A. Noushin , Exact particle flow for nonlinear filters , in Signal processing, sensor fusion, and target recognition XIX, vol. 7697, SPIE, 2010, pp. 92–110

    2010

  32. [40]

    F. E. De Melo, S. Maskell, M. F asiolo, and F. Daum , Stochastic particle flow for nonlinear high-dimensional filtering problems , arXiv preprint arXiv:1511.01448, (2015)

    2015 arXiv

  33. [41]

    de Wiljes, S

    J. de Wiljes, S. Reich, and W. Stannat, Long-time stability and accuracy of the ensemble Kalman– Bucy filter for fully observed processes and small measurement noise , SIAM Journal on Applied Dynamical Systems, 17 (2018), pp. 1152–1181

    2018

  34. [42]

    Del Moral and A

    P. Del Moral and A. Guionnet, On the stability of interacting processes with applications to filtering and genetic algorithms, in Annales de l’Institut Henri Poincar´ e (B) Probability and Statistics, vol. 37, Elsevier, 2001, pp. 155–194

    2001

  35. [43]

    Del Moral, A

    P. Del Moral, A. Kurtzmann, and J. Tugaut, On the stability and the uniform propagation of chaos of a class of extended ensemble Kalman–Bucy filters , SIAM Journal on Control and Optimization, 55 (2017), pp. 119–155

    2017

  36. [44]

    Del Moral and J

    P. Del Moral and J. Tugaut , On the stability and the uniform propagation of chaos properties of ensemble Kalman–Bucy filters , Ann. Appl. Probab., 28 (2018), pp. 790–850

    2018

  37. [45]

    R. A. DeVore and G. G. Lorentz, Constructive approximation, vol. 303, Springer Science & Business Media, 1993

    1993

  38. [46]

    Divol, J

    V. Divol, J. Niles-Weed, and A.-A. Pooladian , Optimal transport map estimation in general function spaces, The Annals of Statistics, 53 (2025), pp. 963–988

    2025

  39. [47]

    Doucet and A

    A. Doucet and A. M. Johansen , A tutorial on particle filtering and smoothing: Fifteen years later , Handbook of nonlinear filtering, 12 (2009), pp. 656–704

    2009

  40. [48]

    S. S. Dragomir , On discrete generalization of Pachpatte’s inequality and applications , Bulletin math´ ematique de la Soci´ et´ e des Sciences Math´ ematiques de Roumanie, 36 (1992), pp. 45–58

    1992

  41. [49]

    Drugowitsch, G

    J. Drugowitsch, G. C. DeAngelis, E. M. Klier, D. E. Angelaki, and A. Pouget , Optimal multisensory decision-making in a reaction-time task , Elife, 3 (2014), p. e03005

    2014

  42. [50]

    R. M. Dudley, The speed of mean Glivenko-Cantelli convergence , The Annals of Mathematical Statis- tics, 40 (1969), pp. 40–50

    1969

  43. [51]

    R. M. Dudley , Universal Donsker classes and metric entropy , in Selected Works of RM Dudley, Springer, 2010, pp. 345–365

    2010

  44. [52]

    T. A. El Moselhy and Y. M. Marzouk , Bayesian inference with optimal maps , Journal of Compu- tational Physics, 231 (2012), pp. 7815–7850

    2012

  45. [53]

    G. Evensen, Sequential data assimilation with a nonlinear quasi-geostrophic model using Monte Carlo methods to forecast error statistics, Journal of Geophysical Research: Oceans, 99 (1994), pp. 10143– 10162

    1994

  46. [54]

    Evensen, Data Assimilation: The Ensemble Kalman Filter , vol

    G. Evensen, Data Assimilation: The Ensemble Kalman Filter , vol. 2, Springer, 2009

    2009

  47. [55]

    Fenchel, On conjugate convex functions , Canadian Journal of Mathematics, 1 (1949), pp

    W. Fenchel, On conjugate convex functions , Canadian Journal of Mathematics, 1 (1949), pp. 73 – 77

    1949

  48. [56]

    Gin´e and R

    E. Gin´e and R. Nickl, Mathematical foundations of infinite-dimensional statistical models , Cambridge university press, 2021

    2021

  49. [57]

    Gonz ´alez-Sanz and S

    A. Gonz ´alez-Sanz and S. Sheng , Linearization of Monge-Amp´ ere equations and statistical applica- tions, arXiv preprint arXiv:2408.06534, (2024)

    2024

  50. [58]

    N. J. Gordon, D. J. Salmond, and A. F. Smith, Novel approach to nonlinear/non-Gaussian Bayesian state estimation, in IEE Proceedings F-radar and signal processing, vol. 140, IET, 1993, pp. 107–113

    1993

  51. [59]

    Grange, M

    D. Grange, M. Al-Jarrah, R. Baptista, A. Taghvaei, T. T. Georgiou, S. Phillips, and A. Tan- nenbaum, Computational optimal transport and filtering on Riemannian manifolds , IEEE Control Systems Letters, (2023)

    2023

  52. [60]

    D. G. Hoag, Apollo navigation, guidance, and control systems: a progress report , MIT Instrumentation Laboratory, 1969

    1969

  53. [61]

    Hosseini, A

    B. Hosseini, A. W. Hsu, and A. Taghvaei , Conditional optimal transport on function spaces , SIAM/ASA Journal on Uncertainty Quantification, 13 (2025), pp. 304–338. ERROR ANALYSIS OF TRIANGULAR OPTIMAL TRANSPORT MAPS FOR FILTERING 31

    2025

  54. [62]

    H¨utter and P

    J.-C. H¨utter and P. Rigollet, Minimax rates of estimation for smooth optimal transport maps , The Annals of Statistics, 49 (2021), pp. 1166–1194

    2021

  55. [63]

    N. J. Irons, M. Scetbon, S. Pal, and Z. Harchaoui , Triangular flows for generative modeling: Statistical consistency, smoothness classes, and fast rates , in International Conference on Artificial Intelligence and Statistics, PMLR, 2022, pp. 10161–10195

    2022

  56. [64]

    Javaheri, D

    A. Javaheri, D. Lautier, and A. Galli , Filtering in finance , Wilmott, 3 (2003), pp. 67–83

    2003

  57. [65]

    R. E. Kalman, A new approach to linear filtering and prediction problems, Journal of Basic Engineering, 82 (1960), pp. 35–45

    1960

  58. [66]

    R. E. Kalman and R. S. Bucy , New results in linear filtering and prediction theory , Journal of basic engineering, 83 (1961), pp. 95–108

    1961

  59. [67]

    Kantas, A

    N. Kantas, A. Doucet, S. S. Singh, and J. M. Maciejowski , An overview of sequential Monte Carlo methods for parameter estimation in general state-space models , IF AC Proceedings Volumes, 42 (2009), pp. 774–785

    2009

  60. [68]

    Kelly, K

    D. Kelly, K. J. Law, and A. M. Stuart, Well-posedness and accuracy of the ensemble Kalman filter in discrete and continuous time , Nonlinearity, 27 (2014), p. 2579

    2014

  61. [69]

    J. W. Kim and P. G. Mehta , Duality for nonlinear filtering i: Observability , IEEE Transactions on Automatic Control, (2023)

    2023

  62. [70]

    Koltchinskii and D

    V. Koltchinskii and D. Panchenko, Rademacher processes and bounding the risk of function learning, in High dimensional probability II, Springer, 2000, pp. 443–457

    2000

  63. [71]

    K. P. Kording, J. B. Tenenbaum, and R. Shadmehr , The dynamics of memory as a consequence of optimal adaptation to a changing body , Nature neuroscience, 10 (2007), pp. 779–786

    2007

  64. [72]

    Korotin, L

    A. Korotin, L. Li, A. Genevay, J. M. Solomon, A. Filippov, and E. Burnaev, Do neural optimal transport solvers work? a continuous Wasserstein-2 benchmark , Advances in neural information processing systems, 34 (2021), pp. 14593–14605

    2021

  65. [73]

    Kutschireiter, S

    A. Kutschireiter, S. C. Surace, and J.-P. Pfister , The Hitchhiker’s guide to nonlinear filtering , Journal of Mathematical Psychology, 94 (2020), p. 102307

    2020

  66. [74]

    Kwiatkowski and J

    E. Kwiatkowski and J. Mandel , Convergence of the square root ensemble Kalman filter in the large ensemble limit , SIAM/ASA Journal on Uncertainty Quantification, 3 (2015), pp. 1–17

    2015

  67. [75]

    Le Gland, V

    F. Le Gland, V. Monbet, and V. Tran , Large sample asymptotics for the ensemble Kalman filter , PhD thesis, INRIA, 2009

    2009

  68. [76]

    L. J. Levy , The Kalman filter: navigation ’s integration workhorse , GPS World, 8 (1997), pp. 65–71

    1997

  69. [77]

    Makkuva, A

    A. Makkuva, A. Taghvaei, S. Oh, and J. Lee , Optimal transport mapping via input convex neural networks, in International Conference on Machine Learning, PMLR, 2020, pp. 6672–6681

    2020

  70. [78]

    Manarvi and T

    A. Manarvi and T. Henderson , Application of Kalman filters in orbit determination: A literature survey, European Space Agency, Contributions, Attachments, (2023)

    2023

  71. [79]

    Mandel, L

    J. Mandel, L. Cobb, and J. D. Beezley , On the convergence of the ensemble Kalman filter , Appli- cations of Mathematics, 56 (2011), pp. 533–541

    2011

  72. [80]

    Marzouk, T

    Y. Marzouk, T. Moselhy, M. Parno, and A. Spantini , Sampling via measure transport: An intro- duction, in Handbook of Uncertainty Quantification , Springer, (2016), pp. 1–41

    2016

  73. [81]

    Marzouk, Z

    Y. Marzouk, Z. R. Ren, S. W ang, and J. Zech , Distribution learning via neural differential equa- tions: a nonparametric statistical perspective , Journal of Machine Learning Research, 25 (2024), pp. 1–61

    2024

  74. [82]

    D. A. Mesa, J. Tantiongloc, M. Mendoza, S. Kim, and T. P. Coleman , A distributed framework for the construction of transport maps , Neural computation, 31 (2019), pp. 613–652

    2019

  75. [83]

    S. P. Meyn and R. L. Tweedie , Markov chains and stochastic stability , Springer Science & Business Media, 2012

    2012

  76. [84]

    Pachpatte, On discrete inequalities related to Gronwall’s inequality , in Proceedings of the Indian Academy of Sciences-Section A, vol

    B. Pachpatte, On discrete inequalities related to Gronwall’s inequality , in Proceedings of the Indian Academy of Sciences-Section A, vol. 85, Springer, 1977, pp. 26–40

    1977

  77. [85]

    G. A. Padmanabha and N. Zabaras , Solving inverse problems using conditional invertible neural networks, Journal of Computational Physics, 433 (2021), p. 110194

    2021

  78. [86]

    Park and K

    J. Park and K. Muandet, Towards empirical process theory for vector-valued functions: Metric entropy of smooth function classes , in International Conference on Algorithmic Learning Theory, PMLR, 2023, pp. 1216–1260

    2023

  79. [87]

    Pathiraja, S

    S. Pathiraja, S. Reich, and W. Stannat, Mckean–Vlasov SDEs in nonlinear filtering , SIAM Journal 32 M. AL-JARRAH, B. HOSSEINI, N. JIN, M. MARTINO, A. TAGHVAEI on Control and Optimization, 59 (2021), pp. 4188–4215

    2021

  80. [88]

    Peyr´e, M

    G. Peyr´e, M. Cuturi, et al., Computational optimal transport, Foundations and Trends® in Machine Learning, 11 (2019), pp. 355–607

    2019

  81. [89]

    D. Ray, J. Murgoitio-Esandi, A. Dasgupta, and A. A. Oberai , Solution of physics-based inverse problems using conditional generative adversarial networks with full gradient penalty , Computer Methods in Applied Mechanics and Engineering, 417 (2023), p. 116338

    2023

  82. [90]

    D. Ray, H. Ramaswamy, D. V. Patel, and A. A. Oberai , The efficacy and generalizability of conditional GANs for posterior inference in physics-based inverse problems , Numerical Algebra, Control and Optimization, (2022)

    2022

  83. [91]

    Rebeschini, Nonlinear filtering in high dimension , PhD thesis, Princeton University, 2014

    P. Rebeschini, Nonlinear filtering in high dimension , PhD thesis, Princeton University, 2014

    2014

  84. [92]

    Rebeschini and R

    P. Rebeschini and R. V an Handel, Can local particle filters beat the curse of dimensionality? , The Annals of Applied Probability, 25 (2015), pp. 2809–2866

    2015

  85. [93]

    Reich, A dynamical systems framework for intermittent data assimilation , BIT Numerical Analysis, 51 (2011), pp

    S. Reich, A dynamical systems framework for intermittent data assimilation , BIT Numerical Analysis, 51 (2011), pp. 235–249

    2011

  86. [94]

    Reich, A nonparametric ensemble transform method for Bayesian inference , SIAM Journal on Sci- entific Computing, 35 (2013), pp

    S. Reich, A nonparametric ensemble transform method for Bayesian inference , SIAM Journal on Sci- entific Computing, 35 (2013), pp. A2013–A2024

    2013

  87. [95]

    Reich, Data assimilation: The Schr¨ odinger perspective, Acta Numerica, 28 (2019), pp

    S. Reich, Data assimilation: The Schr¨ odinger perspective, Acta Numerica, 28 (2019), pp. 635–711

    2019

  88. [96]

    Reich and C

    S. Reich and C. Cotter , Probabilistic forecasting and Bayesian data assimilation , Cambridge Uni- versity Press, 2015

    2015

  89. [97]

    Ristic, S

    B. Ristic, S. Arulampalam, and N. Gordon , Beyond the Kalman filter: Particle filters for tracking applications, Artech house, 2003

    2003

  90. [98]

    S ¨arkk¨a and L

    S. S ¨arkk¨a and L. Svensson , Bayesian filtering and smoothing , vol. 17, Cambridge university press, 2023

    2023

  91. [99]

    Saumard and J

    A. Saumard and J. A. Wellner , Log-concavity and strong log-concavity: a review, Statistics surveys, 8 (2014), p. 45

    2014

  92. [100]

    Sever, S

    S. Sever, S. Dragomir, et al. , Some Gronwall type inequalities and applications , Nova Science Pub Inc, (2003)

    2003

  93. [101]

    Shalev-Shwartz and S

    S. Shalev-Shwartz and S. Ben-David, Understanding machine learning: From theory to algorithms , Cambridge university press, 2014

    2014

  94. [102]

    Spantini, R

    A. Spantini, R. Baptista, and Y. Marzouk , Coupling techniques for nonlinear ensemble filtering , SIAM Review, 64 (2022), pp. 921–953

    2022

  95. [103]

    Sznitman , Topics in propagation of chaos , in Ecole d’´ et´ e de probabilit´ es de Saint-Flour XIX—1989, Springer, 2006, pp

    A.-S. Sznitman , Topics in propagation of chaos , in Ecole d’´ et´ e de probabilit´ es de Saint-Flour XIX—1989, Springer, 2006, pp. 165–251

    1989

  96. [104]

    Taghvaei and B

    A. Taghvaei and B. Hosseini, An optimal transport formulation of Bayes’ law for nonlinear filtering algorithms, IEEE 61st Conference on Decision and Control (CDC), (2022), pp. 6608–6613

    2022

  97. [105]

    Taghvaei and P

    A. Taghvaei and P. G. Mehta, Optimal transportation methods in nonlinear filtering , IEEE Control Systems Magazine, 41 (2021), pp. 34–49

    2021

  98. [106]

    Taghvaei and P

    A. Taghvaei and P. G. Mehta , A survey of feedback particle filter and related controlled interacting particle systems (CIPS) , Annual Reviews in Control, (2023)

    2023

  99. [107]

    X. T. Tong, A. J. Majda, and D. Kelly, Nonlinear stability and ergodicity of ensemble based Kalman filters, Nonlinearity, 29 (2016), p. 657

    2016

  100. [108]

    Truccolo, U

    W. Truccolo, U. T. Eden, M. R. Fellows, J. P. Donoghue, and E. N. Brown , A point process framework for relating neural spiking activity to spiking history, neural ensemble, and extrinsic covariate effects, Journal of neurophysiology, 93 (2005), pp. 1074–1089

    2005

  101. [109]

    V an De Geer, A new approach to least-squares estimation, with applications , The Annals of Statis- tics, (1987), pp

    S. V an De Geer, A new approach to least-squares estimation, with applications , The Annals of Statis- tics, (1987), pp. 587–602

    1987

  102. [110]

    van de Geer, M-estimation using penalties or sieves , Journal of Statistical Planning and Inference, 108 (2002), pp

    S. van de Geer, M-estimation using penalties or sieves , Journal of Statistical Planning and Inference, 108 (2002), pp. 55–69

    2002

  103. [111]

    A. W. van der V aart and J. A. Wellner, Weak Convergence and Empirical Processes—With Ap- plications to Statistics , Springer, 1996

    1996

  104. [112]

    V an Handel, Observability and nonlinear filtering , Probability theory and related fields, 145 (2009), pp

    R. V an Handel, Observability and nonlinear filtering , Probability theory and related fields, 145 (2009), pp. 35–74

    2009

  105. [113]

    R. V an Handel, Nonlinear filtering and systems theory , in Proceedings of the 19th International Sym- posium on Mathematical Theory of Networks and Systems (MTNS semi-plenary paper), 2010. ERROR ANALYSIS OF TRIANGULAR OPTIMAL TRANSPORT MAPS FOR FILTERING 33

    2010

  106. [114]

    P. J. V an Leeuwen, Nonlinear data assimilation in geosciences: an extremely efficient particle filter , Quarterly Journal of the Royal Meteorological Society, 136 (2010), pp. 1991–1999

    2010

  107. [115]

    Vershynin, High-dimensional probability: An introduction with applications in data science , Cam- bridge university press, 2018

    R. Vershynin, High-dimensional probability: An introduction with applications in data science , Cam- bridge university press, 2018

    2018

  108. [116]

    Villani, Optimal Transport: Old and New , vol

    C. Villani, Optimal Transport: Old and New , vol. 338, Springer, 2009

    2009

  109. [117]

    M. J. W ainwright, High-dimensional statistics: A non-asymptotic viewpoint , vol. 48, Cambridge uni- versity press, 2019

    2019

  110. [118]

    W ang and Y

    S. W ang and Y. Marzouk , On minimax density estimation via measure transport , arXiv preprint arXiv:2207.10231, (2022)

    2022 arXiv

  111. [119]

    Z. O. W ang, R. Baptista, Y. Marzouk, L. Ruthotto, and D. Verma , Efficient neural network approaches for conditional optimal transport with applications in Bayesian inference , SIAM Journal on Scientific Computing, 47 (2025), pp. C979–C1005

    2025

  112. [120]

    T. Yang, R. S. Laugesen, P. G. Mehta, and S. P. Meyn , Multivariable feedback particle filter , Automatica, 71 (2016), pp. 10–23

    2016

  113. [121]

    T. Yang, P. G. Mehta, and S. P. Meyn , A mean-field control-oriented approach to particle filtering , in Proceedings of the 2011 American Control Conference, IEEE, 2011, pp. 2037–2043

    2011

  114. [122]

    sup ϕ∈F 1 N NX i=1 ϕ(xi) − µ(ϕ) # ≤ 2 N E

    Z. Yang, C. Jin, Z. W ang, M. W ang, and M. I. Jordan, On function approximation in reinforcement learning: Optimism in the face of large state spaces , 2020. Supplementary Materials A. Examples of stable divergences. Below we list a few commonly used divergences D which are s...

    2020

  115. [123]

    There exists a universal constant Cexp > 0 such that E " sup ϕ∈F 1 N NX i=1 ϕ(xi) − µ(ϕ) # ≤ CexpJN (F ) .(B.8)

  116. [124]

    There exists a universal constant Cprob > 0 such that for any t ≥ 0, (B.9) P " sup ϕ∈F 1 N NX i=1 ϕ(xi) − µ(ϕ) ≥ Cprob JN (F ) + r r t N + R t N !# ≤ exp(−t). Proof. The bound in (B.8) follows from [46, Prop. A.2] by taking, in the notation of that result, ϵ = eϵ = η = 0. We n...

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.