An efficient algorithm for entropic optimal transport under martingale-type constraints

Holakou Rahmanian; Lexing Ying; Michael Shavlovsky; Tesi Xiao; Xun Tang

REVIEW 3 major objections 3 minor 57 references

An efficient algorithm for entropic optimal transport under martingale-type constraints

T0 review · 3 major / 3 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 This paper claims that entropic optimal transport under martingale-type constraints, including discrete martingale optimal transport, is solved efficiently by Sinkhorn-type iterations with sparse Newton steps, achieving super-exponential…

desk verdict Plausible algorithmic extension for discrete MOT (Sinkhorn + sparse Newton), but the central convergence/accuracy trade-off is only claimed 'in practice' and the body was unreadable; still worth sending to a serious referee. read the letter →

arxiv 2508.17641 v1 pith:4TIRLT65 submitted 2025-08-25 math.OC cs.NAmath.NA

Xun Tang , Michael Shavlovsky , Holakou Rahmanian , Tesi Xiao , Lexing Ying This is my paper · ORCID

classification math.OCcs.NAmath.NA MSC 49Q2290C25

keywords entropicoptimaltransportmartingaleSinkhornalgorithmsparseNewtonmethoddualobjectiveHessianstructuralconstraintsapproximateconstraintsatisfactionsuper-exponentialconvergence

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

This paper is trying to establish that entropic optimal transport under martingale-type constraints, including discrete martingale optimal transport, can be solved efficiently by Sinkhorn-type iterations accelerated with sparse Newton steps. The key claim is that these iterations converge super-exponentially in practice while keeping the total violation of the martingale constraints below a user-controllable threshold. Because exact martingale conditions are typically infeasible, the paper deliberately solves an entropically regularized problem that returns an approximately constraint-satisfied coupling. If correct, this gives a practical, scalable solver for a broad class of structurally constrained transport problems, not only the martingale case.

What carries the argument

The central object is the entropically regularized dual problem for martingale-type optimal transport, and the mechanism is the sparse Newton iteration applied to that dual. At each step, instead of building and solving a dense Newton system for the dual variables, the algorithm uses the approximate sparsity of the Hessian $H$ of the dual objective, so the linear solve is restricted to a sparse system. The Sinkhorn-type (Bregman) updates handle the entropic and marginal structure, while the Newton corrections handle the row-wise (in)equality constraints. The approximate sparsity is what converts an otherwise costly second-order method into one whose per-iteration cost stays manageable.

What would settle it

Take a discrete martingale OT instance with a dense, nonlocal cost matrix and measure the density of the dual Hessian at iteration one; if the Hessian is not approximately sparse, or if a truncated sparse Newton step fails to decrease the dual objective, the central convergence claim fails. Alternatively, run the algorithm to its stopping threshold and compute the total martingale constraint violation on a fine grid of regularization strengths; if the violation cannot be made to track the prescribed threshold, the controllability claim fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is that the dual objective of the entropically regularized martingale optimal transport problem has an approximately sparse Hessian, and that this sparsity can be exploited inside Sinkhorn-type iterations to obtain a fast algorithm. The authors formulate discrete martingale optimal transport as an entropic OT problem whose (super-)martingale conditions become row-wise equality or inequality constraints on the coupling matrix, then derive Sinkhorn-type updates combined with sparse Newton steps on the dual. They report that the resulting algorithms show super-exponential convergence and robustness, with the total constraint violation controllable through the chosen threshold. The claim is that the approach extends to the prevalent class of OT problems with structural row-wise constraints.

Load-bearing premise

The method works only if the Hessian of the dual objective is approximately sparse in the regimes that matter, and only if an entropically regularized solution that approximately satisfies the martingale constraints is acceptable.

Editorial extensions

If this is right

Discrete martingale optimal transport becomes solvable by an iterative entropic method with second-order acceleration, so larger instances than those reachable by dense Newton or plain Sinkhorn become practical.
Any optimal transport problem whose structural constraints are row-wise equalities or inequalities on the coupling matrix falls under the same algorithm, broadening the class of solvable structured OT problems.
The entropic formulation gives the user a controllable trade-off: smaller regularization improves constraint satisfaction but makes the problem harder, and the threshold on total constraint violations can be set in advance.
If the observed super-exponential convergence holds beyond the tested regimes, high-accuracy solutions require only a few outer iterations, making the method competitive with specialized MOT solvers.

Reading between the lines

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

The practical payoff depends on how often the dual Hessian is genuinely sparse in realistic instances; for low-temperature (small $\epsilon$) entropic OT, the kernel becomes nearly rank-one and the Hessian may concentrate, which could help or hurt depending on the cost structure.
Because the method returns an approximate solution with controlled constraint violation, applications that need exact martingale pricing may need a post-projection step; the paper does not claim exact constraint satisfaction.
A natural testable extension is to compare the sparse-Newton Sinkhorn method against a primal-dual first-order solver on high-dimensional marginals, measuring both total violation and wall-clock time; this would separate the benefit of second-order acceleration from the benefit of entropic smoothing.

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.

Desk Editor's Note

read the letter

Here's the short version: the abstract describes a legitimate algorithmic idea—entropic regularization of discrete MOT, with Sinkhorn-type iterations accelerated by sparse Newton steps on the dual Hessian. The martingale conditions as row-wise equality constraints on the coupling matrix is a clean observation that naturally covers a broader class of structural OT problems. The authors are also candid that exact martingale conditions are typically infeasible, so they aim for approximately constrained solutions. That is honest and reasonable.

The soft spot is exactly where the stress-test note lands. The abstract says the algorithms enjoy super-exponential convergence and controllable constraint violations 'in practice.' No theorem is claimed, but no experiments or quantitative bounds are visible either. The load-bearing question is whether there is a regime where the entropic regularization parameter is small enough that the constraint violations fall under the advertised tolerance, while the Hessian remains sparse and well-conditioned enough for quadratic Newton convergence to hold. The abstract doesn't address this trade-off. I cannot say the approach fails—the logic is plausible—but the evidence presented here does not demonstrate the central claim. The phrase 'in practice' is doing a lot of work, and with the full text corrupted on my end, I could not check the actual tables, plots, or proofs.

I would not flag this as circular or fitting; it is a forward optimization method and the convergence claim is empirical rather than derived from the desired output. The novelty is incremental rather than paradigm-shifting, but the combination for MOT with sparse Hessians appears to be a genuine extension of existing Sinkhorn/Newton work, and the authors don't oversell it in the abstract.

My recommendation: send it to peer review. The claims are substantive, the problem is timely, and a referee with a readable manuscript can test the central trade-off—either by checking the numerical experiments or trying to construct a counterexample. A desk rejection based on the abstract alone would be premature, especially since the corrupted text prevents anyone from fairly evaluating the full contribution. In a reading group, I'd bring it up for discussion of the regularization-versus-sparsity trade-off, but I wouldn't rely on the results until they're independently verifiable.

Referee Report

3 major / 3 minor

Summary. The manuscript proposes entropic regularization for discrete martingale optimal transport and for the broader class of optimal transport problems with row-wise linear (super-/sub-)martingale-type constraints. It presents Sinkhorn-type algorithms with sparse Newton iterations that exploit the approximate sparsity of the Hessian of the dual objective, and it claims that these algorithms achieve super-exponential convergence in practice while keeping total constraint violations below controllable thresholds. The abstract is readable, but the supplied full text is not: the body is character-corrupted and effectively unreadable, so no derivations, algorithm pseudocode, convergence statements, or experimental tables can be verified. My report is therefore necessarily based on the abstract and on the structure of the claims.

Significance. If the claims could be verified, the paper would address a real computational need: discrete martingale optimal transport is a structurally constrained OT problem with applications in quantitative finance, and an efficient solver with explicit control over constraint violations would be valuable. The idea of combining entropic regularization with sparse Newton iterations is plausible and potentially novel. However, the significance cannot be assessed from the submitted text: the central claims are stated as empirical observations ('in practice') with no supporting theorem, no quantitative bound linking the regularization parameter to constraint violation, and no readable numerical evidence. No machine-checked proofs, reproducible code, or falsifiable predictions are visible. The contribution is therefore currently unsubstantiated.

major comments (3)

[Abstract] The central claim that the proposed algorithms 'in practice' enjoy both super-exponential convergence and robustness with controllable thresholds for total constraint violations is not supported by any equation, theorem, or experiment in the readable portion of the manuscript. Since the full text is unreadable, there is no way to check whether a convergence proof or numerical verification exists. This is load-bearing: the paper's stated contribution rests on these two properties, and neither is presently evidenced.
[Abstract] The manuscript concedes that exact martingale conditions are 'typically infeasible' and that an entropically regularized solution only approximately satisfies the constraints, but it provides no quantitative trade-off between the regularization parameter, the amplitude of total constraint violations, and the sparsity or conditioning of the Hessian. Without such a bound, the claim of 'controllable thresholds' is not established, and it remains possible that the algorithm is fast only in regimes where the constraint violation exceeds any meaningful tolerance. The authors should state an explicit relationship between the regularization parameter and the constraint-violation norm, and should demonstrate empirically that the Hessian remains sufficiently sparse and well conditioned as the regularization parameter is decreased.
[Full text] The body of the manuscript is not readable; it appears as a character-encoding corruption rather than as coherent text. Consequently, none of the algorithm definitions, assumptions, derivations, or numerical results can be checked. This is not a typographical or stylistic issue but a fundamental obstacle to review. A resubmission must contain a readable full text, including clear statements of the algorithm, the convergence result (with proof or precise reference), and experimental protocols and results.

minor comments (3)

[Full text] The text contains a reference to 'arXiv:2508.17638v1 [cs.CV]', which appears to belong to a different paper and should be removed or corrected.
[Abstract] The term 'super-exponential convergence' is used without a formal definition; the authors should specify whether they claim local quadratic convergence, an iteration-complexity bound, or an empirical rate.
[Abstract] The phrase 'controllable thresholds for total constraint violations' should be made precise: the norm in which the violation is measured and the sense in which the threshold is controllable (e.g., via the regularization parameter or via a posteriori bounds) are not stated.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found in the available text

full rationale

The available evidence, chiefly the abstract, describes a forward optimization method: an entropic regularized formulation for martingale-constrained optimal transport, solved by Sinkhorn-type iterations with sparse Newton steps, with convergence and constraint-violation behavior reported 'in practice.' No quantity is fitted to data and then renamed as a prediction, no definition is circularly expressed in terms of the target output, and no load-bearing self-citation chain is visible. The abstract's 'controllable thresholds' claim is empirical rather than derived from the desired output, so even if it is unsupported, that is a correctness or evidence concern, not circularity. The full-text rendering is corrupted and unreadable, preventing any equation-level check of the derivation chain, but the instructions require quoting a specific reduction before claiming circularity, and no such reduction can be exhibited from the available text. Accordingly, the honest finding is no significant circularity.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

No new physical or mathematical entities are introduced; the paper contributes an algorithm. The key assumptions are the row-wise constraint representation, Hessian sparsity, and the acceptability of approximate constraint satisfaction.

free parameters (2)

Entropic regularization parameter
Controls the trade-off between smoothness and exact constraint satisfaction; abstract says regularization is used because exact martingale conditions are infeasible. Not fitted, but chosen by user.
Sparsity threshold for approximate Hessian
The method relies on approximate sparsity of the Hessian; the threshold that determines which entries are kept is a hand-set algorithmic parameter.

assumptions (3)

domain assumption Martingale and super-martingale conditions are equivalent to row-wise equality and inequality constraints on the coupling matrix.
Stated in abstract as the basis for applying the method to a class of structural OT problems. Standard in discrete MOT.
domain assumption The Hessian of the dual objective is approximately sparse.
The algorithm's efficiency depends on this sparsity; it is asserted as an empirical property, not proven in the abstract.
domain assumption Exact martingale conditions are infeasible, so an entropically regularized approximate solution is acceptable.
Shapes the problem definition and the meaning of the output.

how reviews work

0 comments

Cite this review

Pith. "Pith review of An efficient algorithm for entropic optimal transport under martingale-type constraints." pith.science (2026). https://pith.science/paper/4TIRLT65

@misc{pith2026250817641,
  author       = {Pith},
  title        = {Pith review of: An efficient algorithm for entropic optimal transport under martingale-type constraints},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4TIRLT65}},
  note         = {Machine review of arXiv:2508.17641}
}

read the original abstract

This work introduces novel computational methods for entropic optimal transport (OT) problems under martingale-type conditions. The considered problems include the discrete martingale optimal transport (MOT) problem. Moreover, as the (super-)martingale conditions are equivalent to row-wise (in-)equality constraints on the coupling matrix, our work applies to a prevalent class of OT problems with structural constraints. Inspired by the recent empirical success of Sinkhorn-type algorithms, we propose an entropic formulation for the MOT problem and introduce Sinkhorn-type algorithms with sparse Newton iterations that utilize the (approximate) sparsity of the Hessian matrix of the dual objective. As exact martingale conditions are typically infeasible, we adopt entropic regularization to find an approximate constraint-satisfied solution. We show that, in practice, the proposed algorithms enjoy both super-exponential convergence and robustness with controllable thresholds for total constraint violations.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 50 canonical work pages

[1]

, Cuturi , M

barticle Peyr \'e , G. , Cuturi , M. , : Computational optimal transport: With applications to data science . Foundations and Trends in Machine Learning 11 ( 5-6 ), 355 -- 607 ( 2019 ) barticle

work page 2019
[2]

, Touzi , N

barticle Tan , X. , Touzi , N. : Optimal transportation under controlled stochastic dynamics . The Annals of Probability 41 ( 5 ), 3201 -- 3240 ( 2013 ) barticle

work page 2013
[3]

, Henry-Labordere , P

barticle Beiglb \"o ck , M. , Henry-Labordere , P. , Penkner , F. : Model-independent bounds for option prices—a mass transport approach . Finance and Stochastics 17 , 477 -- 501 ( 2013 ) barticle

work page 2013
[4]

, Henry-Labord \`e re , P

barticle Galichon , A. , Henry-Labord \`e re , P. , Touzi , N. : A stochastic control approach to no-arbitrage bounds given marginals, with an application to lookback options . The Annals of Applied Probability 24 ( 1 ), 312 -- 336 ( 2014 ) barticle

work page 2014
[5]

, Soner , H.M

barticle Dolinsky , Y. , Soner , H.M. : Martingale optimal transport and robust hedging in continuous time . Probability Theory and Related Fields 160 ( 1-2 ), 391 -- 427 ( 2014 ) barticle

work page 2014
[6]

, Ob \'o j , J

barticle Guo , G. , Ob \'o j , J. : Computational methods for martingale optimal transport problems . The Annals of Applied Probability 29 ( 6 ), 3311 -- 3347 ( 2019 ) barticle

work page 2019
[7]

, Stebegg , F

barticle Nutz , M. , Stebegg , F. : Canonical supermartingale couplings . The Annals of Probability 46 ( 6 ), 3351 -- 3398 ( 2018 ) barticle

work page 2018
[8]

, Kim , Y.-H

barticle Ghoussoub , N. , Kim , Y.-H. , Lim , T. : Structure of optimal martingale transport plans in general dimensions . The Annals of Probability 47 ( 1 ), 109 -- 164 ( 2019 ) barticle

work page 2019

Show all 57 references

[9]

, Trevisan , D

botherref Huesmann , M. , Trevisan , D. : A benamou--brenier formulation of martingale optimal transport (2019) botherref

2019
[10]

, Corbetta , J

barticle Alfonsi , A. , Corbetta , J. , Jourdain , B. : Sampling of probability measures in the convex order by Wasserstein projection . Annales de l'Institut Henri Poincar \'e (B) Probabilit \'e s et Statistiques 56 ( 3 ), 1706 -- 1729 ( 2020 ) 10.1214/19-AIHP1014 barticle

2020 doi
[11]

, Pammer , G

barticle Backhoff-Veraguas , J. , Pammer , G. : Stability of martingale optimal transport and weak optimal transport . The Annals of Applied Probability 32 ( 1 ), 721 -- 752 ( 2022 ) barticle

2022
[12]

: Continuity of the martingale optimal transport problem on the real line

barticle Wiesel , J. : Continuity of the martingale optimal transport problem on the real line . The Annals of Applied Probability 33 ( 6A ), 4645 -- 4692 ( 2023 ) barticle

2023
[13]

: From (martingale) schrodinger bridges to a new class of stochastic volatility model

botherref Henry-Labordere , P. : From (martingale) schrodinger bridges to a new class of stochastic volatility model. Available at SSRN 3353270 (2019) botherref

2019
[14]

, Wiesel , J

botherref Nutz , M. , Wiesel , J. : On the martingale schr " odinger bridge between two distributions. arXiv preprint arXiv:2401.05209 (2024) botherref

2024 arXiv
[15]

: Optimal Transport Methods in Economics

bbook Galichon , A. : Optimal Transport Methods in Economics . Princeton University Press , ( 2018 ) bbook

2018
[16]

, Morstatter , F

barticle Mehrabi , N. , Morstatter , F. , Saxena , N. , Lerman , K. , Galstyan , A. : A survey on bias and fairness in machine learning . ACM computing surveys (CSUR) 54 ( 6 ), 1 -- 35 ( 2021 ) barticle

2021
[17]

, Hardt , M

bbook Barocas , S. , Hardt , M. , Narayanan , A. : Fairness and Machine Learning: Limitations and Opportunities . MIT press , ( 2023 ) bbook

2023
[18]

, Murthy , K

bchapter Si , N. , Murthy , K. , Blanchet , J. , Nguyen , V.A. : Testing group fairness via optimal transport projections . In: International Conference on Machine Learning , pp. 9649 -- 9659 ( 2021 ). PMLR bchapter

2021
[19]

, De Bie , T

barticle Buyl , M. , De Bie , T. : Optimal transport of classifiers to fairness . Advances in Neural Information Processing Systems 35 , 33728 -- 33740 ( 2022 ) barticle

2022
[20]

: On the methods of measuring association between two attributes

barticle Yule , G.U. : On the methods of measuring association between two attributes . Journal of the Royal Statistical Society 75 ( 6 ), 579 -- 652 ( 1912 ) barticle

1912
[21]

: A relationship between arbitrary positive matrices and doubly stochastic matrices

barticle Sinkhorn , R. : A relationship between arbitrary positive matrices and doubly stochastic matrices . The annals of mathematical statistics 35 ( 2 ), 876 -- 879 ( 1964 ) barticle

1964
[22]

: Sinkhorn distances: Lightspeed computation of optimal transport

botherref Cuturi , M. : Sinkhorn distances: Lightspeed computation of optimal transport. Advances in neural information processing systems 26 (2013) botherref

2013
[23]

, Carlier , G

barticle Benamou , J.-D. , Carlier , G. , Cuturi , M. , Nenna , L. , Peyr \'e , G. : Iterative bregman projections for regularized transportation problems . SIAM Journal on Scientific Computing 37 ( 2 ), 1111 -- 1138 ( 2015 ) barticle

2015
[24]

, Rahmanian , H

botherref Tang , X. , Rahmanian , H. , Shavlovsky , M. , Thekumparampil , K.K. , Xiao , T. , Ying , L. : A sinkhorn-type algorithm for constrained optimal transport. arXiv preprint arXiv:2403.05054 (2024) botherref

2024 arXiv
[25]

: An unconstrained convex programming view of linear programming

barticle Fang , S.-C. : An unconstrained convex programming view of linear programming . Zeitschrift f \"u r Operations Research 36 , 149 -- 161 ( 1992 ) barticle

1992
[26]

: An explicit analysis of the entropic penalty in linear programming

bchapter Weed , J. : An explicit analysis of the entropic penalty in linear programming . In: Conference On Learning Theory , pp. 1841 -- 1855 ( 2018 ). PMLR bchapter

2018
[27]

, Guo , G

barticle Eckstein , S. , Guo , G. , Lim , T. , Ob \'o j , J. : Robust pricing and hedging of options on multiple assets and its numerics . SIAM Journal on Financial Mathematics 12 ( 1 ), 158 -- 188 ( 2021 ) barticle

2021
[28]

, Stebegg , F

barticle Nutz , M. , Stebegg , F. , Tan , X. : Multiperiod martingale transport . Stochastic Processes and their Applications 130 ( 3 ), 1568 -- 1615 ( 2020 ) barticle

2020
[29]

, Ho , N

barticle Lin , T. , Ho , N. , Cuturi , M. , Jordan , M.I. : On the complexity of approximating multimarginal optimal transport . Journal of Machine Learning Research 23 ( 65 ), 1 -- 43 ( 2022 ) barticle

2022
[30]

, Alaya , M.Z

barticle Chapel , L. , Alaya , M.Z. , Gasso , G. : Partial optimal tranport with applications on positive-unlabeled learning . Advances in Neural Information Processing Systems 33 , 2903 -- 2913 ( 2020 ) barticle

2020
[31]

, Nguyen , H

bchapter Le , K. , Nguyen , H. , Nguyen , K. , Pham , T. , Ho , N. : On multimarginal partial optimal transport: Equivalent forms and computational complexity . In: International Conference on Artificial Intelligence and Statistics , pp. 4397 -- 4413 ( 2022 ). PMLR bchapter

2022
[32]

, Nguyen , D

bchapter Nguyen , K. , Nguyen , D. , Pham , T. , Ho , N. , : Improving mini-batch optimal transport via partial transportation . In: International Conference on Machine Learning , pp. 16656 -- 16690 ( 2022 ). PMLR bchapter

2022
[33]

, Nguyen , T.D

bchapter Nguyen , A.D. , Nguyen , T.D. , Nguyen , Q.M. , Nguyen , H.H. , Nguyen , L.M. , Toh , K.-C. : On partial optimal transport: Revising the infeasibility of sinkhorn and efficient gradient methods . In: Proceedings of the AAAI Conference on Artificial Intelligence , vol....

2024
[34]

, Gasnikov , A

bchapter Dvurechensky , P. , Gasnikov , A. , Kroshnin , A. : Computational optimal transport: Complexity by accelerated gradient descent is better than by sinkhorn’s algorithm . In: International Conference on Machine Learning , pp. 1367 -- 1376 ( 2018 ). PMLR bchapter

2018
[35]

, Ho , N

bchapter Lin , T. , Ho , N. , Jordan , M. : On efficient optimal transport: An analysis of greedy and accelerated mirror descent algorithms . In: International Conference on Machine Learning , pp. 3982 -- 3991 ( 2019 ). PMLR bchapter

2019
[36]

, Jepson , A.D

botherref Kemertas , M. , Jepson , A.D. , Farahmand , A.-m. : Efficient and accurate optimal transport with mirror descent and conjugate gradients. arXiv preprint arXiv:2307.08507 (2023) botherref

2023 arXiv
[37]

, Shavlovsky , M

botherref Tang , X. , Shavlovsky , M. , Rahmanian , H. , Tardini , E. , Thekumparampil , K.K. , Xiao , T. , Ying , L. : Accelerating sinkhorn algorithm with sparse newton iterations. To appear in ICLR 2024 (2024) https://arxiv.org/abs/2401.12253 arXiv:2401.12253 [math.OC] botherref

2024 arXiv
[38]

, Chang , Y

bchapter Dong , A. , Chang , Y. , Zheng , Z. , Mishne , G. , Bai , J. , Zhang , R. , Buchner , K. , Liao , C. , Diaz , F. : Towards recency ranking in web search . In: Proceedings of the Third ACM International Conference on Web Search and Data Mining , pp. 11 -- 20 ( 2010 ) bchapter

2010
[39]

, Shokouhi , M

bchapter Dai , N. , Shokouhi , M. , Davison , B.D. : Learning to rank for freshness and relevance . In: Proceedings of the 34th International ACM SIGIR Conference on Research and Development in Information Retrieval , pp. 95 -- 104 ( 2011 ) bchapter

2011
[40]

, Garakani , A.B

botherref Momma , M. , Garakani , A.B. , Sun , Y. : Multi-objective relevance ranking (2019) botherref

2019
[41]

, Haramaty , E

bchapter Carmel , D. , Haramaty , E. , Lazerson , A. , Lewin-Eytan , L. : Multi-objective ranking optimization for product search using stochastic label aggregation . In: Proceedings of The Web Conference 2020 , pp. 373 -- 383 ( 2020 ) bchapter

2020
[42]

, : Learning to rank for information retrieval

barticle Liu , T.-Y. , : Learning to rank for information retrieval . Foundations and Trends in Information Retrieval 3 ( 3 ), 225 -- 331 ( 2009 ) barticle

2009
[43]

: An Introduction to Information Retrieval

bbook Manning , C.D. : An Introduction to Information Retrieval . Cambridge university press , ( 2009 ) bbook

2009
[44]

: Tres observaciones sobre el algebra lineal

barticle Birkhoff , G. : Tres observaciones sobre el algebra lineal . Univ. Nac. Tucuman, Ser. A 5 , 147 -- 154 ( 1946 ) barticle

1946
[45]

, : Optimal Transport: Old and New vol

bbook Villani , C. , : Optimal Transport: Old and New vol. 338 . Springer , ??? ( 2009 ) bbook

2009
[46]

, Pandey , R

bchapter McAuley , J. , Pandey , R. , Leskovec , J. : Inferring networks of substitutable and complementary products . In: Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pp. 785 -- 794 ( 2015 ) bchapter

2015
[47]

, Juillet , N

barticle Br \"u ckerhoff , M. , Juillet , N. : Instability of martingale optimal transport in dimension d >= 2 . Electronic Communications in Probability 27 , 1 -- 10 ( 2022 ) barticle

2022
[48]

: All of Nonparametric Statistics

bbook Wasserman , L. : All of Nonparametric Statistics . Springer , ( 2006 ) bbook

2006
[49]

, Bach , F

botherref Weed , J. , Bach , F. : Sharp asymptotic and finite-sample rates of convergence of empirical measures in wasserstein distance (2019) botherref

2019
[50]

, Niles-Weed , J

botherref Chewi , S. , Niles-Weed , J. , Rigollet , P. : Statistical optimal transport. arXiv preprint arXiv:2407.18163 (2024) botherref

2024 arXiv
[51]

, Vandenberghe , L

bbook Boyd , S.P. , Vandenberghe , L. : Convex Optimization . Cambridge university press , ( 2004 ) bbook

2004
[52]

, Wright , S.J

bbook Nocedal , J. , Wright , S.J. : Numerical Optimization . Springer , ( 1999 ) bbook

1999
[53]

, Ye , Y

bbook Luenberger , D.G. , Ye , Y. , : Linear and Nonlinear Programming vol. 2 . Springer , ( 1984 ) bbook

1984
[54]

, Neuberger , A

barticle Hobson , D. , Neuberger , A. : Robust bounds for forward start options . Mathematical Finance: An International Journal of Mathematics, Statistics and Financial Economics 22 ( 1 ), 31 -- 56 ( 2012 ) barticle

2012
[55]

: The (2) limit in the random assignment problem

barticle Aldous , D.J. : The (2) limit in the random assignment problem . Random Structures & Algorithms 18 ( 4 ), 381 -- 418 ( 2001 ) barticle

2001
[56]

a rvelin , K. , Kek \

barticle J \"a rvelin , K. , Kek \"a l \"a inen , J. : Cumulated gain-based evaluation of ir techniques . ACM Transactions on Information Systems (TOIS) 20 ( 4 ), 422 -- 446 ( 2002 ) barticle

2002
[57]

write newline

" write newline "" before.all 'output.state := FUNCTION string.to.integer 't := t text.length 'k := #1 'char.num := t char.num #1 substring 's := s is.num s "." = or char.num k = not and char.num #1 + 'char.num := while char.num #1 - 'char.num := t #1 char.num substring FUNCTI...

Pith tools

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

[1] [1]

, Cuturi , M

barticle Peyr \'e , G. , Cuturi , M. , : Computational optimal transport: With applications to data science . Foundations and Trends in Machine Learning 11 ( 5-6 ), 355 -- 607 ( 2019 ) barticle

work page 2019

[2] [2]

, Touzi , N

barticle Tan , X. , Touzi , N. : Optimal transportation under controlled stochastic dynamics . The Annals of Probability 41 ( 5 ), 3201 -- 3240 ( 2013 ) barticle

work page 2013

[3] [3]

, Henry-Labordere , P

barticle Beiglb \"o ck , M. , Henry-Labordere , P. , Penkner , F. : Model-independent bounds for option prices—a mass transport approach . Finance and Stochastics 17 , 477 -- 501 ( 2013 ) barticle

work page 2013

[4] [4]

, Henry-Labord \`e re , P

barticle Galichon , A. , Henry-Labord \`e re , P. , Touzi , N. : A stochastic control approach to no-arbitrage bounds given marginals, with an application to lookback options . The Annals of Applied Probability 24 ( 1 ), 312 -- 336 ( 2014 ) barticle

work page 2014

[5] [5]

, Soner , H.M

barticle Dolinsky , Y. , Soner , H.M. : Martingale optimal transport and robust hedging in continuous time . Probability Theory and Related Fields 160 ( 1-2 ), 391 -- 427 ( 2014 ) barticle

work page 2014

[6] [6]

, Ob \'o j , J

barticle Guo , G. , Ob \'o j , J. : Computational methods for martingale optimal transport problems . The Annals of Applied Probability 29 ( 6 ), 3311 -- 3347 ( 2019 ) barticle

work page 2019

[7] [7]

, Stebegg , F

barticle Nutz , M. , Stebegg , F. : Canonical supermartingale couplings . The Annals of Probability 46 ( 6 ), 3351 -- 3398 ( 2018 ) barticle

work page 2018

[8] [8]

, Kim , Y.-H

barticle Ghoussoub , N. , Kim , Y.-H. , Lim , T. : Structure of optimal martingale transport plans in general dimensions . The Annals of Probability 47 ( 1 ), 109 -- 164 ( 2019 ) barticle

work page 2019