Pith. sign in

REVIEW 2 major objections 6 minor 58 references

Sinkhorn Linearization and the Spectral Proxy: Unifying the Statistical and Algorithmic Theory of Feature-Parameterized Inverse Optimal Transport via a Single Spectral Sandwich

T0 review · 2 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read One spectral bound on the Sinkhorn linearization drives identifiability, stability, convergence, and sparse recovery in inverse optimal transport.

desk verdict A mostly sound spectral framework for inverse OT, with a clean core bound driving T1/T3/T4, but the sparsistency theorem rests on an openly admitted and empirically unsatisfied irrepresentability bridge. read the letter →

arxiv 2608.13201 v1 pith:DAVKSLYP submitted 2026-08-13 stat.ML cs.LGmath.OCmath.STstat.TH

Han Dong , Jiaming Li , Yongqiang Gong , Ruixi Li , Yin Liu This is my paper · ORCID
classification stat.MLcs.LGmath.OCmath.STstat.TH MSC 49Q2262F1262J0790C25
keywords inverseoptimaltransportSinkhornlinearizationspectralproxyentropicregularizationidentifiabilitysparsistencystrongmonotonicitymisspecifiedestimation
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 sets out to show that feature-parameterized inverse optimal transport—recovering the cost parameters of an entropic transport problem from observed transition probabilities—is a well-posed statistical problem rather than a heuristic inversion. Its key move is to differentiate the Sinkhorn map with respect to the cost and prove a two-sided spectral bound on the restricted Hessian: the inverse of the Hessian is sandwiched between $(\pi_{\min}/\varepsilon)I$ and $(\pi_{\max}/\varepsilon)I$. From this one sandwich the paper derives four theorems and an observation: global identifiability up to a gauge kernel, support recovery for $\ell_1$-penalized estimation, strong monotonicity of the feature-moment map with a Lipschitz inverse, monotone gradient-descent convergence, and convergence of misspecified estimators to the projection of the truth onto the OT model set. If the claims hold, practitioners get explicit dimension limits, Lipschitz constants, strong-convexity moduli, and exponential rate constants for a class of problems that previously lacked them.

What carries the argument

The Sinkhorn linearization is the implicit-function derivative of the entropic plan with respect to the cost, $\delta x = -B H_T^{-1} B^\top \delta c$, obtained by differentiating the KKT conditions of the entropy-regularized problem. Its spectral proxy, $\delta x_{\mathrm{SSP}} = -(1/\varepsilon) P_T D_\pi P_T \delta c$, replaces the dense inverse by two tangent-space projections surrounding elementwise multiplication by the plan entries, preserving all spectral bounds. The load-bearing identity is the spectral sandwich $(\pi_{\min}/\varepsilon)I \preceq H_T^{-1} \preceq (\pi_{\max}/\varepsilon)I$: because every plan entry lies between $\pi_{\min}$ and $\pi_{\max}$, the inverse restricted Hessian inherits uniform spectral control, and that control propagates directly to the singular-value lower bound that all downstream theorems use.

What would settle it

At a dense grid of parameter values with fixed positive marginals and $\lambda_{\min}(\Sigma)>0$, compute the exact Jacobian $J_\theta$ by automatic differentiation and test whether $\sigma_{\min}(J_\theta) \ge (\pi_{\min}(\theta)/(a_{\max}\varepsilon))\sqrt{\lambda_{\min}(\Sigma)}$; any single violation would falsify Proposition 3.3 and, with it, the common core of T1, T3, and T4.

Watch

Extended reading notes

Core claim

The central discovery is that the statistical and algorithmic theory of feature-parameterized inverse optimal transport radiates from a single spectral bound. The paper proves that the restricted Hessian of the entropic OT objective, pulled back to the tangent space of feasible plans, has inverse sandwiched between $(\pi_{\min}/\varepsilon)I$ and $(\pi_{\max}/\varepsilon)I$; consequently the Jacobian of the conditional transition operator with respect to the cost parameters obeys $\sigma_{\min}(J_\theta) \ge (\pi_{\min}(\theta)/(a_{\max}\varepsilon))\sqrt{\lambda_{\min}(\Sigma)}$ at every $\theta$ with positive plan entries and positive $\lambda_{\min}(\Sigma)$. This bound yields global injectivity on the quotient by the gauge kernel (T1), strong monotonicity and a Lipschitz inverse for the feature-moment map (T3), local strong convexity of the cross-entropy objective with monotone gradient descent (T4), and support recovery for the $\ell_1$-penalized estimator under an additional irrepresentability condition (T2); the misspecification analysis (O5) identifies the estimator's limit as the projection of the truth onto the OT model set.

Load-bearing premise

The load-bearing premise is that the information matrix at the true parameter satisfies an irrepresentability condition, a requirement the paper cannot check a priori and that its own 40-seed experiment finds violated in every trial.

Editorial extensions

If this is right

  • On the quotient space $\mathbb{R}^F/N_\Phi$, the map from cost parameters to conditional transition operators is globally injective; on the original space it is injective exactly when $\mathrm{rank}(\Sigma)=F$, which forces the feature-dimension cap $F\le(K-1)^2$.
  • On any compact convex parameter domain, the inverse map from observed transition operators to cost parameters is Lipschitz with constant at most $\varepsilon\|\Phi^\top S_a\|_{\mathrm{op}}/(\pi_{\min}\lambda_{\min}(\Sigma))$, so small observation noise cannot produce arbitrarily large parameter error.
  • Near the true parameter, the population cross-entropy is strongly convex with modulus at least $\pi_{\min}^2\lambda_{\min}(\Sigma)/\varepsilon^2$, so fixed-step-size gradient descent converges monotonically from any initialization inside its basin.
  • Under irrepresentability of the information matrix and all-coordinate score concentration, the $\ell_1$-penalized estimator recovers the true support with failure probability at most $C_1\exp(-2n t_n^2/\Delta_{\max}^2)$, and the debiased refit on the recovered support is asymptotically normal.
  • When the data are not generated by the OT model, the estimator converges to the pre-image (under the T3 inverse map) of the cross-entropy projection of the truth onto the OT model set, rather than to any true parameter.

Reading between the lines

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

  • Because the core bound degrades as $\pi_{\min}\to 0$, a practical design rule the paper leaves implicit is that $\varepsilon$ should be chosen to keep $\pi_{\min}(\theta,\varepsilon)\sqrt{\lambda_{\min}(\Sigma)}/\varepsilon$ bounded away from zero, rather than merely set small.
  • The paper's own 40-seed check finds the irrepresentability condition violated in every trial, which suggests that plain $\ell_1$ penalization in raw feature coordinates may not be the right sparse-recovery device for generic feature parameterizations; a preconditioned or debiased estimator whose effective information matrix inherits Gram-matrix irrepresentability is a natural testable alternative.
  • The spectral proxy's geometric form—project, weight by plan entries, project back—predicts that inverse-problem conditioning is controlled by where the plan mass sits; one could test this by placing features on low-mass regions of the plan and measuring the condition number of the sensitivity Gram matrix.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 6 minor

Summary. This paper develops a frequentist statistical theory for inverse entropic optimal transport (IOT) under feature-parameterized costs C_theta(i,j)=-theta^T phi(i,j), with observations given by conditional transition operators. The main technical object is the Sinkhorn linearization, the exact sensitivity of the entropic OT plan to the cost, whose tangent-space restricted Hessian satisfies the spectral sandwich (10). From this sandwich the authors derive a core singular-value lower bound (Proposition 3.3) and use it to drive four theorems and one observation: T1 (global identifiability on a gauge quotient with dimension bound F <= (K-1)^2), T2 (l1-support recovery under a Hessian irrepresentability condition), T3 (strong monotonicity and Lipschitz stability of the feature-moment map), T4 (local strong convexity and monotone gradient-descent convergence), and O5 (convergence to the pseudo-true projection under misspecification). The numerical section reports experiments E1-E10, including support-recovery phase transitions, irrepresentability diagnostics, perturbation transfer, initialization comparisons, and empirical scaling of pi_min with epsilon.

Significance. If the results are taken as stated, the paper makes a useful contribution: it reduces identifiability, well-posedness, and convergence constants to a single spectral bound sigma_min(J_theta) >= (pi_min/(a_max epsilon)) sqrt(lambda_min(Sigma)), with the Sinkhorn linearization as the underlying calculus. The derivation of Proposition 3.3 and the arguments for T1, T3(a), and T4 appear internally sound, and the paper is commendably explicit about which claims are conditional, conjectural, or empirical (e.g., the Holder exponent and the pi_min scaling law). The main caveat is that T2's irrepresentability hypothesis is stated for the OT information matrix H = grad^2 ell(theta*), whereas the only a priori sufficient condition supplied (Lemma 5.2) concerns the feature Gram matrix Sigma, and the transfer bridge is explicitly not assumed. The experiments find the H-irrepresentability condition satisfied 0/40 times, so the sparse-recovery theorem currently has no demonstrated non-vacuous instantiation in the feature-parameterized setting. This limits, but does not destroy, the claimed unification.

major comments (2)
  1. [Section 5.1, Lemma 5.2 and Assumption A2.2; Section 9.2, E8b] The central issue is the gap between Assumption A2.2 and its only a priori sufficient condition. A2.2 requires irrepresentability of H = grad^2 ell(theta*), while Lemma 5.2 verifies irrepresentability of the feature Gram matrix Sigma; the lemma's own note states that transferring Sigma-irrepresentability to H 'requires an additional bridge' that is 'not separately assumed in this paper.' Since no such bridge is proved and the numerical check E8b reports the H-condition satisfied 0/40 times (with values in [1.36, 7.43]), Theorem 5.1 is currently a conditional Lasso-type statement rather than an established consequence of the spectral framework. I recommend either proving a transfer result (e.g., a quantitative comparison between H and a constant multiple of Sigma under explicit conditions on pi_min, epsilon, and feature normalization) or revising the abstract, introduction, and conclusion so that T2 is presented as conditional on an unverified structural condition instead of as one of the theorems established on the core bound.
  2. [Section 5.1, Theorem 5.1 and Assumption A2.1] The global selection condition is not formalized as a precise hypothesis. The proof's primal-dual witness construction establishes that the constructed restricted candidate satisfies the KKT conditions; equality with the global l1-penalized estimator requires the additional condition, stated inside A2.1, that the neighborhood U contains both the PDW candidate and the global solution and that no lower minimum exists outside U. This condition is only illustrated by examples ('for example, the penalized objective is convex...'), and no sufficient condition is established in the IOT setting. As written, Theorem 5.1 therefore proves support recovery for the PDW local solution, not necessarily for the global estimator btheta_lasso_n. The hypothesis must be stated formally if the theorem is to support the claimed conclusion about the global l1 estimator.
minor comments (6)
  1. [Section 3.2] The sentence '(PAP)|_T^{-1} != P A^{-1}P' uses the unqualified inverse of a singular operator; please write the inverse on the range explicitly to avoid ambiguity.
  2. [Section 3.5 and throughout] The 'core spectral bound' is referred to as (3.3) in many places, but no displayed equation with that number appears; renumber the displays or fix the cross-references.
  3. [Abstract and Definition 3.2] The abstract says the spectral proxy is 'spectrally exact (it preserves all singular-value bounds)', but Definition 3.2 only establishes equality of the two-sided operator-norm envelope; suggest 'preserves the spectral bounds' for accuracy.
  4. [Section 9.6] The subsection title 'The epsilon-scaling law of pi_min' overstates the status of a finite-window log-log fit; the surrounding text correctly labels alpha_eff as an empirical exponent, so consider retitling the subsection.
  5. [Section 5.1 and Section 9.2, E9] The text in Section 5.1 mentions a (6x-156x) comparison between the range-based Hoeffding constant and the Bernstein constant, while E9 reports ratios in [13.7, 33.9]; reconcile the two ranges or clarify that they refer to different thresholds and settings.
  6. [Section 9, Experimental protocol] The implementation-specific path '09 iot theory/output/experiments e1 e10.json' should be replaced by a stable artifact reference or removed from the final manuscript.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the core spectral bound is derived from the exact Sinkhorn linearization, and the paper's documented gaps are incompleteness rather than circular reductions.

full rationale

The main derivation chain is self-contained. Lemma 3.1 obtains the spectral sandwich (10) directly from H_T = ε B^T D_π^{-1} B and the entrywise bounds on 1/π_ij, and Proposition 3.3 propagates this sandwich through the exact linearization δx = -B H_T^{-1} B^T δc rather than through the spectral proxy (13), so the core bound is not defined into existence. T1's global injectivity is proved by a direct three-step KKT/composition argument that does not depend on the spectral bound for its conclusion. T3(a) and T4 follow from standard monotone-operator and Hessian arguments once the pointwise spectral bound is uniformized on compact sets; no fitted constants enter those proofs. T2 is explicitly conditional on Assumptions A2.1–A2.3, and the paper honestly discloses the limitation that Lemma 5.2 supplies only Gram-matrix irrepresentability while the transfer to the OT information matrix H 'requires an additional bridge' that 'is not separately assumed in this paper'; E8b then reports the H-level irrepresentability condition satisfied 0/40 times. That is a scope gap, not a circular step, because the theorem is an if-then statement under its assumptions. O5's residual bound is labeled 'definitional' and bears no derivational weight, and the empirical exponents α_eff are explicitly described as finite-window fits rather than theoretical predictions. There is no load-bearing self-citation chain: the uniqueness, sparsistency, monotone-operator, and misspecified M-estimation results cited are external classical references. The paper's dependency table somewhat overstates T1's reliance on (3.3), but that rhetorical emphasis does not make any theorem reduce to its own inputs.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The theoretical claims rest on standard regularity assumptions: positive marginals, rank of the gauge-cleaned Gram, compactness for global statements, and the usual Lasso sufficient conditions. The only fitted numbers are empirical exponents reported in experiments, clearly labeled and not used in the derivations. No invented entities are postulated.

free parameters (2)
  • alpha_eff (empirical Holder exponent estimates) = 0.353, 0.266, 0.369, 0.298 across four settings; 0.048 after cleanup at epsilon=0.1
    Fitted by log-log regression of parameter displacement vs perturbation; explicitly labeled as setting-dependent and not a universal constant. Not used in any proof.
  • pi_min scaling exponent (E10) = ~19.9
    Empirical fit of pi_min >= c epsilon^alpha over a finite window; the paper notes exponential decay can mimic a large power exponent. Not used in proofs.
assumptions (6)
  • domain assumption Assumption 4.2: all marginal entries strictly positive and full support
    Needed for the plan to be strictly positive, for the tangent-space dimension (K-1)^2, and for pi -> Q to be injective. Entered at Section 4.1.
  • domain assumption Assumption 4.3: rank(Sigma)=F (features linearly independent off the gauge)
    Needed for original-space identifiability and for the core spectral bound to be non-vacuous. Section 4.1.
  • domain assumption Assumption 6.6: compact convex parameter domain with inf pi_min > 0
    Needed for the global strong monotonicity in T3(b); the paper states local T3(a) does not need it. Section 6.
  • domain assumption A2.1 local RSC, A2.2 irrepresentability of H, A2.3 beta-min and penalty window, plus a global selection condition
    Standard Lasso sufficient conditions imported for T2; the paper notes the global selection condition is additional. Section 5.1.
  • domain assumption Unique pseudo-true point and compact theta for O5
    Used for the misspecified M-estimation consistency argument. Section 8.
  • standard math Standard large-deviation and monotone-operator results (Hoeffding, matrix Hoeffding, Lasso theory of [30,26], Zarantonello-Minty [29,17])
    Used for concentration, sparsistency, and the Lipschitz inverse of monotone maps. Sections 5, 6, 7.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Sinkhorn Linearization and the Spectral Proxy: Unifying the Statistical and Algorithmic Theory of Feature-Parameterized Inverse Optimal Transport via a Single Spectral Sandwich." pith.science (2026). https://pith.science/paper/DAVKSLYP

@misc{pith2026260813201,
  author       = {Pith},
  title        = {Pith review of: Sinkhorn Linearization and the Spectral Proxy: Unifying the Statistical and Algorithmic Theory of Feature-Parameterized Inverse Optimal Transport via a Single Spectral Sandwich},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DAVKSLYP}},
  note         = {Machine review of arXiv:2608.13201}
}
read the original abstract

We develop the statistical and algorithmic theory of inverse optimal transport (IOT) under the feature-parameterized cost C_theta(i,j) = -theta^T phi(i,j). The core technical contribution is the Sinkhorn linearization -- the implicit-function sensitivity of the entropic OT plan to the cost -- together with its spectral proxy, a formula that is spectrally exact yet geometrically transparent. The restricted Hessian on the tangent space satisfies the spectral sandwich (pi_min/epsilon) I <= H_T^{-1} <= (pi_max/epsilon) I, yielding the single core bound sigma_min >= (pi_min/(a_max epsilon)) sqrt(lambda_min(Sigma)) that drives the entire theory. On this core we establish four theorems and one observation. T1 (identifiability): theta is globally injective on the quotient of the gauge kernel, with dimension bound F <= (K-1)^2. T2 (sparsistency): the l1-penalized estimator recovers the true support under irrepresentability and score concentration, with exponential failure probability. T3 (well-posedness): the feature-moment map M(theta) = Phi^T x_theta is strongly monotone, and the inverse is Lipschitz with constant L <= epsilon ||Phi^T S_a||_op / (pi_min lambda_min(Sigma)). T4 (convergence): local strong convexity with mu >= pi_min^2 lambda_min(Sigma) / epsilon^2 guarantees monotone gradient descent convergence. O5 (misspecification): the estimator converges to the OT-model projection of the truth; the Holder continuity of the projection map is assessed numerically, yielding setting-dependent empirical exponents alpha_eff in (0,1).

Figures

Figures reproduced from arXiv: 2608.13201 by the authors.

Figure 1
Figure 1. Condition-number divergence as ε → 0 [PITH_FULL_IMAGE:figures/full_fig_p023_1.png] view at source ↗
Figure 2
Figure 2. Generality verification 23 [PITH_FULL_IMAGE:figures/full_fig_p023_2.png] view at source ↗
Figure 3
Figure 3. Theorem-compatible sparsistency diagnostics: (a) sample-complexity phase transi [PITH_FULL_IMAGE:figures/full_fig_p024_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Effect of the voting mechanism on the recovery probability [PITH_FULL_IMAGE:figures/full_fig_p025_4.png]
Figure 5
Figure 5. Figure 5: Numerical verification of the irrepresentability condition [PITH_FULL_IMAGE:figures/full_fig_p025_5.png]
Figure 6
Figure 6. Figure 6: Comparison of the Bernstein and Hoeffding constants [PITH_FULL_IMAGE:figures/full_fig_p026_6.png]
Figure 7
Figure 7. Figure 7: The ε ′ -bias V-shape 9.4 T4: Convergence (E4) Comparison of initialization schemes (E4, 40 restarts). The success rates of the three schemes—random / multiscale (coarse warm-start) / prior-guided (initialized on the true support)— are 0.25, 0.95, 0.20 respectively, wi…
Figure 8
Figure 8. Figure 8: Comparison of initialization schemes 9.5 O5: Misspecification (E5) Under a non-OT generating mechanism (row softmax plus noise), the projection residual is 4.50, larger than the true-OT residual 2.43 but much smaller than the random-model residual 10.63 (ratio 0.42); t…
Figure 9
Figure 9. Figure 9: Verification of the H¨older exponent across settings [PITH_FULL_IMAGE:figures/full_fig_p029_9.png]
Figure 10
Figure 10. Figure 10: Adam optimization residual analysis at ε = 0.1: the CE has flattened after 500 iterations while the parameters continue to drift; the responses at the three smallest perturbation amplitudes are submerged by the optimization residual 29 [PITH_FULL_IMAGE:figures/full_f…
Figure 11
Figure 11. Figure 11: The ε-scaling law of πmin 10 Conclusion This paper establishes the statistical and algorithmic theory of IOT under a feature-parameterized cost. The central technical contribution is the Sinkhorn linearization and the spectral proxy (Section 3): the spectral sandwich …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 49 canonical work pages

  1. [2]

    Bao et al

    Z. Bao et al. , Well-posedness of the inverse bregman optimal transport problem , (2025). Working paper

    work page 2025

  2. [4]

    Bernton, P

    E. Bernton, P. Ghosal, and M. Nutz , Stability of entropic optimal transport and schr \"o dinger bridges , Journal of Functional Analysis, 283 (2022), p. 109622

    work page 2022

  3. [5]

    Bigot, E

    J. Bigot, E. Cazelles, and N. Papadakis , Central limit theorems for entropy-regularized optimal transport on finite spaces and statistical applications , Electronic Journal of Statistics, 13 (2019), pp. 5112--5150

    work page 2019

  4. [6]

    Cominetti and J

    R. Cominetti and J. San Mart \'i n , Asymptotic analysis of the exponential penalty trajectory in linear programming , Mathematical Programming, 67 (1994), pp. 169--187

    work page 1994

  5. [7]

    Cuturi , Sinkhorn distances: Lightspeed computation of optimal transport , Advances in Neural Information Processing Systems, 26 (2013)

    M. Cuturi , Sinkhorn distances: Lightspeed computation of optimal transport , Advances in Neural Information Processing Systems, 26 (2013)

    work page 2013

  6. [8]

    Dupuy and A

    A. Dupuy and A. Galichon , Personality traits and the marriage market , Journal of Political Economy, 122 (2014), pp. 1271--1319

    work page 2014

  7. [9]

    Galichon , Optimal Transport Methods in Economics , Princeton University Press, 2018

    A. Galichon , Optimal Transport Methods in Economics , Princeton University Press, 2018

    work page 2018

  8. [11]

    Nonlinear Inverse Optimal Transport: Identifiability of the Transport Cost from its Marginals and Optimal Values

    A. Gonz \'a lez-Sanz, M. Groppe, and A. Munk , Nonlinear inverse optimal transport: identifiability of the transport cost from its marginals and optimal values , SIAM Journal on Mathematical Analysis, 56 (2024), pp. 7808--7829. arXiv:2312.05843

    work page Pith review arXiv 2024

Show all 58 references
  1. [12]

    Hallin, E

    M. Hallin, E. del Barrio, J. Cuesta-Albertos, and C. Matr \'a n , Center-outward distribution and quantile functions, annular spreads, and the Mallows--Wasserstein distances between distributions , Annals of Statistics, 49 (2021), pp. 2567--2605

    2021

  2. [14]

    L \'e onard , A survey of the schr \"o dinger problem and some of its connections with optimal transport , Discrete & Continuous Dynamical Systems, 34 (2014), pp

    C. L \'e onard , A survey of the schr \"o dinger problem and some of its connections with optimal transport , Discrete & Continuous Dynamical Systems, 34 (2014), pp. 1533--1574

    2014

  3. [15]

    Mascherpa, A

    M. Mascherpa, A. Ringh, A. Taghvaei, and J. Karlsson , A convex approach for Markov chain estimation from aggregate data via inverse optimal transport , arXiv preprint arXiv:2511.16458, (2025)

    2025

  4. [16]

    Mena and J

    G. Mena and J. Niles-Weed , Statistical bounds for entropic optimal transport: sample complexity and the central limit theorem , in Advances in Neural Information Processing Systems, vol. 32, 2019

    2019

  5. [17]

    G. J. Minty , Monotone (nonlinear) operators in Hilbert space , Duke Mathematical Journal, 29 (1962), pp. 341--346

    1962

  6. [18]

    Persiianov et al

    S. Persiianov et al. , Semi-supervised inverse entropic optimal transport , (2025). Working paper

    2025

  7. [19]

    Peyr \'e and M

    G. Peyr \'e and M. Cuturi , Computational optimal transport , Foundations and Trends in Machine Learning, 11 (2019), pp. 355--607

    2019

  8. [21]

    R. T. Rockafellar , Monotone operators and the proximal point algorithm , SIAM Journal on Control and Optimization, 14 (1976), pp. 877--898

    1976

  9. [22]

    Santambrogio , Optimal Transport for Applied Mathematicians , vol

    F. Santambrogio , Optimal Transport for Applied Mathematicians , vol. 87 of Progress in Nonlinear Differential Equations and their Applications, Birkh \"a user, 2015

    2015

  10. [23]

    Sinkhorn , A relationship between arbitrary positive matrices and doubly stochastic matrices , The Annals of Mathematical Statistics, 35 (1964), pp

    R. Sinkhorn , A relationship between arbitrary positive matrices and doubly stochastic matrices , The Annals of Mathematical Statistics, 35 (1964), pp. 876--879

    1964

  11. [24]

    A. M. Stuart and M.-T. Wolfram , Inverse optimal transport , SIAM Journal on Applied Mathematics, 80 (2020), pp. 599--619

    2020

  12. [25]

    Vacher, B

    A. Vacher, B. Muzellec, A. Nishimura, and A. Dunipace , Bayesian inverse optimal transport , in Proceedings of Machine Learning Research (PMLR), 2021. arXiv:2105.12140

    2021 arXiv

  13. [26]

    M. J. Wainwright , Sharp thresholds for high-dimensional and noisy sparsity recovery using _1 -constrained quadratic programming ( Lasso ) , IEEE Transactions on Information Theory, 55 (2009), pp. 2183--2202

    2009

  14. [27]

    Weed , An explicit analysis of the entropic penalty in linear programming , in Conference on Learning Theory (COLT), 2018

    J. Weed , An explicit analysis of the entropic penalty in linear programming , in Conference on Learning Theory (COLT), 2018

    2018

  15. [28]

    White , Maximum likelihood estimation of misspecified models , Econometrica, 50 (1982), pp

    H. White , Maximum likelihood estimation of misspecified models , Econometrica, 50 (1982), pp. 1--25

    1982

  16. [29]

    E. H. Zarantonello , Solving functional equations by contractive averaging , Tech. Report Technical Report \#160, Mathematics Research Center, U.S. Army, 1960

    1960

  17. [30]

    Zhao and B

    P. Zhao and B. Yu , On model selection consistency of Lasso , Journal of Machine Learning Research, 7 (2006), pp. 2541--2563

    2006

  18. [31]

    2024 , note=

    Identifiability of inverse optimal transport problems on finite spaces , author=. 2024 , note=

    2024

  19. [32]

    SIAM Journal on Mathematical Analysis , volume=

    Nonlinear inverse optimal transport: identifiability of the transport cost from its marginals and optimal values , author=. SIAM Journal on Mathematical Analysis , volume=. 2024 , note=

    2024

  20. [33]

    arXiv preprint arXiv:2410.23146 , year=

    Identifiability and exact reconstruction of the optimal transport cost on finite spaces , author=. arXiv preprint arXiv:2410.23146 , year=

    arXiv

  21. [34]

    2025 , note=

    Well-posedness of the inverse Bregman optimal transport problem , author=. 2025 , note=

    2025

  22. [35]

    arXiv preprint arXiv:2310.05461 , year=

    Sparsistency for inverse optimal transport , author=. arXiv preprint arXiv:2310.05461 , year=

    arXiv

  23. [36]

    2025 , note=

    Semi-supervised inverse entropic optimal transport , author=. 2025 , note=

    2025

  24. [37]

    Advances in Neural Information Processing Systems , volume=

    Sinkhorn distances: Lightspeed computation of optimal transport , author=. Advances in Neural Information Processing Systems , volume=

  25. [38]

    Foundations and Trends in Machine Learning , volume=

    Computational optimal transport , author=. Foundations and Trends in Machine Learning , volume=

  26. [39]

    The Annals of Mathematical Statistics , volume=

    A relationship between arbitrary positive matrices and doubly stochastic matrices , author=. The Annals of Mathematical Statistics , volume=

  27. [40]

    Optimal Transport for Applied Mathematicians , author=

  28. [41]

    A survey of the Schr

    L. A survey of the Schr. Discrete & Continuous Dynamical Systems , volume=

  29. [42]

    Conference on Learning Theory (COLT) , year=

    An explicit analysis of the entropic penalty in linear programming , author=. Conference on Learning Theory (COLT) , year=

  30. [43]

    Mathematical Programming , volume=

    Asymptotic analysis of the exponential penalty trajectory in linear programming , author=. Mathematical Programming , volume=

  31. [44]

    Stability of entropic optimal transport and Schr

    Bernton, Espen and Ghosal, Promit and Nutz, Marcel , journal=. Stability of entropic optimal transport and Schr

  32. [45]

    Electronic Journal of Statistics , volume=

    Central limit theorems for entropy-regularized optimal transport on finite spaces and statistical applications , author=. Electronic Journal of Statistics , volume=

  33. [46]

    Advances in Neural Information Processing Systems , volume=

    Statistical bounds for entropic optimal transport: sample complexity and the central limit theorem , author=. Advances in Neural Information Processing Systems , volume=

  34. [47]

    arXiv preprint arXiv:2109.12004 , year=

    Entropic estimation of optimal transport maps , author=. arXiv preprint arXiv:2109.12004 , year=

    arXiv

  35. [48]

    Optimal Transport Methods in Economics , author=

  36. [49]

    Journal of Political Economy , volume=

    Personality traits and the marriage market , author=. Journal of Political Economy , volume=

  37. [50]

    A convex approach for

    Mascherpa, Michele and Ringh, Axel and Taghvaei, Amirhossein and Karlsson, Johan , journal=. A convex approach for

  38. [51]

    , journal=

    Minty, George J. , journal=. Monotone (nonlinear) operators in

  39. [52]

    SIAM Journal on Control and Optimization , volume=

    Monotone operators and the proximal point algorithm , author=. SIAM Journal on Control and Optimization , volume=

  40. [53]

    Econometrica , volume=

    Maximum likelihood estimation of misspecified models , author=. Econometrica , volume=

  41. [54]

    On model selection consistency of

    Zhao, Peng and Yu, Bin , journal=. On model selection consistency of

  42. [55]

    , journal=

    Wainwright, Martin J. , journal=. Sharp thresholds for high-dimensional and noisy sparsity recovery using _1 -constrained quadratic programming (

  43. [56]

    Proceedings of Machine Learning Research (PMLR) , year=

    Bayesian inverse optimal transport , author=. Proceedings of Machine Learning Research (PMLR) , year=

  44. [57]

    arXiv preprint arXiv:2106.07360 , year=

    Statistical estimation of optimal transport maps: a density estimation approach to learning the optimal transport map between distributions , author=. arXiv preprint arXiv:2106.07360 , year=

    arXiv

  45. [58]

    Center-outward distribution and quantile functions, annular spreads, and the

    Hallin, Marc and del Barrio, Eustasio and Cuesta-Albertos, Juan and Matr. Center-outward distribution and quantile functions, annular spreads, and the. Annals of Statistics , volume=

  46. [59]

    arXiv preprint arXiv:2206.00428 , year=

    Tikhonov regularization of optimal transport mappings , author=. arXiv preprint arXiv:2206.00428 , year=

    arXiv

  47. [60]

    Foundations and Trends in Machine Learning , volume=

    An introduction to matrix concentration inequalities , author=. Foundations and Trends in Machine Learning , volume=

  48. [61]

    SIAM Journal on Applied Mathematics , volume=

    Inverse optimal transport , author=. SIAM Journal on Applied Mathematics , volume=

  49. [62]

    arXiv preprint arXiv:2604.22670 , year=

    Curvature of optimal transport with respect to the cost and applications to inverse optimal transport , author=. arXiv preprint arXiv:2604.22670 , year=

    arXiv

  50. [63]

    Solving functional equations by contractive averaging , author=

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.