A system of Schr\"odinger's problems and functional equations

Jin Feng; Toshio Mikami

REVIEW 1 major objections 4 minor 40 references

A system of Schr\"odinger's problems and functional equations

T0 review · 1 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read An inductively defined chain of Schrödinger-style relative-entropy problems has measurable solutions and unique minimizers at each stage.

desk verdict A genuinely new existence and uniqueness theorem for an iterative Schrödinger system, with a solid proof and an explicit but restrictive convexity assumption; worth refereeing. read the letter →

arxiv 2501.00719 v2 pith:6UPIYVUF submitted 2025-01-01 math.PR math.OC

Toshio Mikami , Jin Feng This is my paper · ORCID

classification math.PRmath.OC MSC 49Q2293E20

keywords Schrödinger'sproblemfunctionalequationKnothe–RosenblattrearrangementstochasticoptimaltransportrelativeentropyBernsteinprocessh-path

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 studies a system of Schrödinger problems: rather than one relative-entropy minimization between two endpoint measures on R^d, the authors minimize a sequence of entropies, one per coordinate block, where each stage's reference measure is the previous stage's optimal coupling tensored with the conditional distribution of the next block under the starting measure μ. The central claim is that this induction is well-posed: at each stage the associated functional equation has a solution, unique up to a multiplicative function of the past coordinates, and, when the entropy is finite, the displayed coupling is the unique minimizer. The result supplies a variational route to a stochastic optimal-transport analog of the Knothe–Rosenblatt rearrangement, an object for which no general existence theorem had been available. A sympathetic reader should care because the construction converts a multi-marginal transport problem into a chain of one-block Schrödinger bridges, each solvable from the previous one.

What carries the argument

The load-bearing object is the inductively defined reference measure π_{0,i} = π_{μ_{i-1},ν_{i-1}} ⊗ μ_{i|i-1} p_i(· | y_{n_{i-1}}) dy-block, together with the functional equation (1.31) whose unknown h_i is the Radon–Nikodym factor making the i-th marginal correct. The proof mechanism is the convexity assumption (A0)(iii): for each new block, y ↦ log p_i(x_{n_i},(y_{n_{i-1}},y)) + ψ_i(y) is convex, which forces the integral operator I_i(φ)(y_{n_{i-1}},·) to be continuous on the interior of its convex domain (Lemma 3.1). Continuity in the new block is then upgraded to measurability in the past coordinates by a selection lemma that chooses a Borel point (y_{n_{i-1}},ξ_i(y_{n_{i-1}})) where the conditional density is positive, allowing the multiplicative ambiguity to be normalized. Finally, uniqueness of the minimizer comes from disintegrating the relative entropy along the past coordinate and invoking the classical I-projection/Schrödinger-bridge uniqueness at almost every slice.

What would settle it

The cleanest falsifier is to exhibit a triple (μ,ν,p) satisfying (A0)–(A1) with V_2 finite for which (1.31) has two solutions not differing by a multiplicative function of y_{n_1}; the theorem's uniqueness assertion would then be false. Numerically, this can be probed by solving the stage-two functional equation by fixed-point iteration on slices and checking whether the assembled coupling reproduces the direct minimizer of V_2.

Watch

Extended reading notes

Core claim

The paper's main theorem (Theorem 2.1) states that under assumptions (A0)–(A1), for each i=2,…,k0, the functional equation (1.31) admits a measurable solution h_i defined on $R^{{n_i}}$, with h_i(y_{n_{i-1}},·) continuous on $R^{{d_i}}$, satisfying (2.9) for fν_{i-1}(y_{n_{i-1}})dy_{n_{i-1}}-almost every past coordinate. The solution is unique up to a multiplicative measurable function of y_{n_{i-1}}, and the measure π_{μ_i,ν_i} defined in (2.13) belongs to the admissible class A(μ_i,ν_i;π_{μ_{i-1},ν_{i-1}}) and is the unique minimizer of V_i whenever V_i is finite. At the base, h1 is a continuous solution of the classical Schrödinger functional equation (1.29), and the first coupling π_{μ_1,ν_1} is the classical Schrödinger bridge. The authors thereby extend Jamison's existence theory for a single bridge to a chain of bridges whose reference measures are built recursively, and they frame the whole system as a stochastic counterpart of the Knothe–Rosenblatt rearrangement.

Load-bearing premise

The whole proof leans on the assumption that, at every stage, the logarithm of the conditional transition density in the new coordinate block plus a fixed continuous function is convex in that block; if that convexity fails, the continuity and measurability arguments used to construct h_i and the minimizer no longer go through.

Editorial extensions

If this is right

At every stage with finite entropy, the inductively built coupling π_{μ_i,ν_i} genuinely belongs to A(μ_i,ν_i;π_{μ_{i-1},ν_{i-1}}), so the chain defines a stochastic analogue of the Knothe–Rosenblatt rearrangement without constructing an explicit triangular map.
When k0=1 the system collapses to the classical Schrödinger problem and its functional equation, recovering the known existence and uniqueness theory as the base case.
For the Gaussian product kernel of Example 2.2, the theorem produces an explicit Bernstein-type probability law on C([0,1];R^2), giving a process-level object from the variational construction.
Because (1.31) is the Euler equation of V_i, the minimizer can be sought by solving one functional equation per coordinate block, so the system suggests a block-by-block Sinkhorn algorithm.
The uniqueness statement fixes the continuation of the chain: once π_{μ_{i-1},ν_{i-1}} is known, the next coupling has no free parameter when V_i is finite.

Reading between the lines

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

A natural test is the zero-noise limit: replacing the transition density by a Brownian kernel with vanishing variance should make the chain of bridges concentrate on the deterministic Knothe–Rosenblatt map, a limit the paper leaves as future work.
Under stronger smoothness or log-concavity of the kernels, the measurable h_i are likely to be continuous in the past coordinates as well, which would resolve the paper's open question about continuous solutions to (1.31).
The recursive construction is not tied to Euclidean space in an essential way; the same block-by-block scheme could be run for any family of positive continuous Markov kernels satisfying the analogous convexity assumption, e.g., on manifolds or graphs.
Computationally, the paper implies a conditional-slice Sinkhorn algorithm: at block i, solve a classical Schrödinger problem for each past coordinate slice and then glue the solutions by the measurable selection; convergence analysis for such an algorithm is not provided here.

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.

Desk Editor's Note

read the letter

Genuinely new and mostly clean. The paper extends Schrödinger's classical problem to an inductive system whose minimizers give a variational, stochastic analogue of the Knothe–Rosenblatt rearrangement. Theorem 2.1 is the real content: existence and uniqueness (up to a multiplicative function of the conditioning variable) for the system of functional equations (1.31), and identification of the unique minimizer of the relative-entropy variational problem (1.27). For k0=1 it reduces to the standard Schrödinger problem.

The proof is solid. The authors use Jamison's theorem for the conditional problems, convexity (Lemma 3.1) to turn (A0)(iii) into continuity of the integral operator, and a selection lemma to handle measurability. I checked the induction and the measurable selection step; it holds together. The assumption (A0)(iii) is strong, but it is explicitly stated and exactly what the proof needs. The finiteness example (Example 2.1) is a nice touch.

Soft spots are minor. Lemma 3.2 is stated without proof, with a note that it follows like Lemma 3.4; I believe that's correct, but in a refereed version they should either prove it or spell out the reduction. The verification that a diffusion transition density satisfies (A0)(iii) is only sketched; for a theorem conditional on (A0), this weakens the sense of applicability, but it doesn't undermine the result. The significance is subfield-level: first existence and uniqueness for a nested Schrödinger system, but not the full stochastic Knothe–Rosenblatt program, which the authors explicitly leave open.

I'd send this to a serious referee. It deserves refereeing rather than desk rejection, and I'd expect acceptance after a modest revision that fills in Lemma 3.2. I would cite it if I worked on multi-marginal entropic transport or triangular transport.

Referee Report

1 major / 4 minor

Summary. The paper introduces an inductive system of variational problems of relative entropy (1.26)-(1.27) with two endpoint marginals, together with an inductively defined system of functional equations (1.29), (1.31), generalizing Schrodinger's problem and functional equation. Under assumptions (A0)-(A1), Theorem 2.1 asserts, for each i=2,...,k0, the existence of a measurable solution h_i of the conditional equation (2.9) with h_i(y_{n_{i-1}},.) continuous, uniqueness up to a multiplicative measurable function of y_{n_{i-1}}, and the unique minimality of the measure pi_{mu_i,nu_i} in (2.13) for V_i whenever V_i is finite. The proof is inductive, using Jamison's theorem for an auxiliary Schrodinger equation, convexity of log p_i + psi_i (Lemma 3.1), weak-continuity and measurability results (Lemmas 3.3-3.6), and a measurable selection argument to regularize h_i.

Significance. The framework is a plausible new route to a stochastic optimal transport analog of the Knothe-Rosenblatt rearrangement. The assumptions are explicit, including the restrictive log-concavity type condition (A0)(iii), and the paper honestly records the limitation that h_i need not be jointly continuous (Remark 2.3). The proof couples standard tools (Jamison's representation, convexity, Lusin's theorem, and a selection lemma) rather than introducing a radically new technique, but the resulting existence and uniqueness theorem for a system of Schrodinger equations appears to be new. I found no circularity or internal inconsistency: the inductive use of previously constructed minimizers is recursion, not a logical loop.

major comments (1)

[Section 3, Lemma 3.2] Lemma 3.2 is stated without proof ('We omit the proof'), but it is used directly in the proof of Theorem 2.1 and in Lemmas 3.3 and 3.4. The indicated argument by analogy with Lemma 3.4 is credible, since the i=1 case can be derived from Lemma 3.1(i) and Theorem 1.1 without Proposition 2.1; however, the manuscript as submitted does not contain that derivation. Because this is a load-bearing step of the induction, please supply a self-contained proof of Lemma 3.2 or an explicit reference to a published theorem that yields continuity of the Schrodinger solution h_1 under (A0)(i)-(ii) and (A1).

minor comments (4)

[Section 4, proof of Proposition 2.1] The assertion that a solution to (2.9) is also a solution to (4.2) is used for the uniqueness claim but is not demonstrated; adding one line with tilde h_i(y_{n_{i-1}},z,y)=h_i(y_{n_{i-1}},y) and integrating z against q_{i-1}dz would make the equivalence explicit.
[Section 4, proof of Theorem 2.1] After (4.8), the measurability of the displayed integral is attributed to Lemma 3.6, but the integrand depends on y_{n_{i-1}} also through xi_i(y_{n_{i-1}}); please state that this dependence can be absorbed into the test function in (3.14).
[Section 3, proof of Lemma 3.6] There are several minor typographical errors, e.g., 'Foy any phi' should be 'For any phi'; a final proofreading pass is needed.
[Remark 2.1(i)] The verification that (A0)(iii) holds for the marginal kernels p_i is only sketched; a sentence noting that log-concavity is preserved under taking marginals (e.g., by Prekopa-Leindler) would make the example more transparent.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the inductive proof builds on Jamison's external existence theorem and explicit convexity assumptions; no step reduces to its own inputs.

full rationale

The central result (Theorem 2.1) is proved by induction. For each i, Proposition 2.1 first solves a conditional Schrödinger functional equation (4.2) by citing Jamison's Theorem 1.1 (an external theorem) for the auxiliary density q_{i-1}, then integrates over z to obtain h_i solving (2.9). This is recursion on previously constructed minimizers, not circularity. The claim that any solution of (2.9) also solves (4.2) is a direct verification from the definitions, and uniqueness up to a multiplicative function of y_{n_{i-1}} follows from Jamison's uniqueness for the auxiliary equation. Lemma 3.4 derives continuity of h_i(y_{n_{i-1}},·) from the explicit convexity assumption (A0)(iii) via Lemma 3.1(ii), with h_i defined as f_{ν_i}/I_i(φ_i); this is an explicit construction, not a hidden use of the conclusion. The measurability step in Theorem 2.1 uses a standard Borel selection lemma ([11]) and Lemma 3.6, whose proof uses the stability result Lemma 3.5 cited from the first author's earlier paper [26]. That citation is load-bearing for the proof but is an external published theorem on continuity of Schrödinger solutions under perturbations, not an assumption equivalent to the target result; hence it does not make the derivation circular. The inductive system (1.26)-(1.27) is defined using previously constructed minimizers, which is legitimate recursion. No fitted parameters are renamed as predictions, and no known result is merely relabeled. The restrictive assumptions (A0)(iii) and the omitted proof of Lemma 3.2 are presentation and scope limitations, not circularity.

Assumptions & free parameters 0 free parameters · 8 assumptions · 0 invented entities

The central claim rests on the structural assumptions (A0) and (A1) on the transition kernel p and the target density fν, plus several external theorems: Jamison's representation and SDE well-posedness, the I-projection minimizer theorem, a stability result for Schrodinger solutions, and a measurable selection lemma. These are stated as background and are not derived in the paper. No numerical parameters are fitted to data.

assumptions (8)

domain assumption Assumption (A0)(i): p ∈ C(R^d × R^d; (0,∞)) with {p(x,·)dy} probability measures, and the integrated kernels p_i are continuous and independent of trailing coordinates.
Defines the transition kernel used to build the relative-entropy costs; the triangular/independence structure is necessary for the inductive definition of the system.
domain assumption Assumption (A0)(ii): there exists ψ1 ∈ C(R^{d1}) such that y ↦ log p1(x_{n1}, y) + ψ1(y) is convex.
Used in Lemma 3.1(i) to guarantee continuity of the integral operator I1 on its domain, a key step in the proof of Lemma 3.2 and the base case.
domain assumption Assumption (A0)(iii): for i=2,...,k0, there exists ψi ∈ C(R^{di}) such that the conditional log-kernel is convex in the y[ni-1+1,ni] block.
This is the load-bearing convexity condition used in Lemma 3.1(ii) to establish continuity of Ii(φi)(yni-1,·), which underpins Lemma 3.4 and the construction of continuous hi.
domain assumption Assumption (A1): ν has a probability density fν and each marginal density fνi is continuous.
Regularity of the target marginal; the openness of fνi^{-1}((0,∞)) (Remark 2.1(iii)) plays a crucial role in the selection lemma and pointwise arguments.
standard math Jamison's representation theorem (Theorem 1.1) and the existence/uniqueness of the h-path SDE (Theorem 1.2).
Provides existence and uniqueness of solutions to the classical Schrodinger functional equation and the Markovian reciprocal process under (H).
standard math I-projection minimizer theorems for Schrodinger problems (Ruschendorf-Thomsen [33]).
Gives that the measure defined via h solves the single-stage variational problem V1 and, in Proposition 2.1, the conditional problem.
standard math Stability of Schrodinger solutions under locally uniform convergence of kernels and weak convergence of marginals (Lemma 3.5, from Mikami [26]).
Used in Lemma 3.6 to transfer weak continuity of the data to weak continuity of the conditional Schrodinger solutions.
standard math Measurable selection lemma (Fleming-Rishel [11], p. 199).
Used in Theorem 2.1 to construct a Borel measurable selector ξi for points where fνi > 0, needed to build a measurable version of hi.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A system of Schr\"odinger's problems and functional equations." pith.science (2026). https://pith.science/paper/6UPIYVUF

@misc{pith2026250100719,
  author       = {Pith},
  title        = {Pith review of: A system of Schr\"odinger's problems and functional equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6UPIYVUF}},
  note         = {Machine review of arXiv:2501.00719}
}

read the original abstract

We propose and study a system of Schr\"odinger's problems and functional equations in probability theory. More precisely, we consider a system of variational problems of relative entropies for probability measures on a Euclidean space with given two endpoint marginals, which can be defined inductively. We also consider an inductively defined system of functional equations, which are Euler's equations for our variational problems. These are generalizations of Schr\"odinger's problem and functional equation. % in probability theory. We prove the existence and uniqueness of solutions to our functional equations, % up to a multiplicative function, from which we show the existence and uniqueness of a minimizer of our variational problem. Our problem gives an approach for a stochastic optimal transport analog of the Knothe--Rosenblatt rearrangement via a variational problem point of view.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 39 canonical work pages

[1]

Beiglb¨ ock, G

M. Beiglb¨ ock, G. Pammer, and A. Posch,The Knothe–Rosenblatt dis- tance and its induced topology , arXiv:2312.16515

arXiv
[2]

Bernstein, Sur les liaisons entre les grandeurs al´ etoires , Verh

S. Bernstein, Sur les liaisons entre les grandeurs al´ etoires , Verh. des intern. Mathematikerkongr. Zurich 1932, Band 1 (1932), pp. 288–309

work page 1932
[3]

V. I. Bogachev, A. V. Kolesnikov, and K. V. Medvedev, On triangular transformations of measures (Russian), Dokl. Akad. Nauk, 396, no. 6 (2004), pp. 727–732

work page 2004
[4]

V. I. Bogachev, A. V. Kolesnikov, and K. V. Medvedev, Triangular transformations of measures (Russian), Mat. Sb., 196, no. 3 (2005), pp. 3–30; translation in Sb. Math., 196, no. 3-4 (2005), pp. 309–335

work page 2005
[5]

Carlier, F

G. Carlier, F. Galichon, and F. Santambrogio, From Knothe’s transport to Brenier’s map and a continuation method for optimal transport , SIAM J. Math. Anal., 41, no. 6 (2010), pp. 2554–2576

work page 2010
[6]

Y. Chen, T. Georgiou, and M. Pavon, Entropic and displacement inter- polation: a computational approach using the Hilbert metric , SIAM J. Appl. Math., 76, no. 6 (2016), pp. 2375–2396

work page 2016
[7]

Y. Chen, T. Georgiou, and M. Pavon, Stochastic control liaisons: Richard Sinkhorn meets Gaspard Monge on a Schr¨ odinger bridge. SIAM Review 63, no. 2, 249–313 (2021)

work page 2021
[8]

Csiszar, I-divergence geometry of probability distributions and mini- mization problems, Ann

I. Csiszar, I-divergence geometry of probability distributions and mini- mization problems, Ann. Probab., 3 (1975), pp. 146–158

work page 1975

Show all 40 references

[9]

Dai Pra, A stochastic control approach to reciprocal diffusion pro- cesses, Appl

P. Dai Pra, A stochastic control approach to reciprocal diffusion pro- cesses, Appl. Math. Optim., 23 (1991), pp. 313–329. 26

1991
[10]

Dall’Aglio, Sugli estremi dei momenti delle funzioni di ripartizione doppia, Ann

G. Dall’Aglio, Sugli estremi dei momenti delle funzioni di ripartizione doppia, Ann. Scuola Norm. Sup. Pisa Cl. Sci. (3), 10, no. 1-2 (1956), pp. 35–74

1956
[11]

W. H. Fleming and R. W. Rishel, Deterministic and Stochastic Optimal Control, Springer, New York, 1975

1975
[12]

W. H. Fleming and H. M. Soner, Controlled Markov Processes and Viscosity Solutions, 2nd ed., Springer-Verlag , New York, 2006

2006
[13]

F¨ ollmer,Random fields and diffusion processes , in ´Ecole d’ ´Et´ e de Probabilit´ es de Saint-Flour XV–XVII, 1985–87, Lecture Notes in Math

H. F¨ ollmer,Random fields and diffusion processes , in ´Ecole d’ ´Et´ e de Probabilit´ es de Saint-Flour XV–XVII, 1985–87, Lecture Notes in Math. 1362, P.L. Hennequin, ed., Springer-Verlag, Berlin, 1988, pp. 101–203

1985
[14]

Fortet, R´ esolution d’un Syst` eme d’´ equations de M

R. Fortet, R´ esolution d’un Syst` eme d’´ equations de M. Schr¨ odinger, J. Math. Pures Appl. (9) 19, 83–105 (1940)

1940
[15]

Ikeda and S

N. Ikeda and S. Watanabe, Stochastic Differential Equations and Dif- fusion Processes, 2nd ed., North–Holland/Kodansha , Tokyo, 2014

2014
[16]

Jamison, Reciprocal processes , Z

B. Jamison, Reciprocal processes , Z. Wahrscheinlichkeitstheorie und Verw. Gebiete, 30 (1974), pp. 65–86

1974
[17]

Jamison, The Markov process of Schr¨ odinger, Z

B. Jamison, The Markov process of Schr¨ odinger, Z. Wahrscheinlichkeit- stheorie und Verw. Gebiete, 32 (1975), pp. 323–331

1975
[18]

Kamijima, A

T. Kamijima, A. Takatsu, K. Funo, and T. Sagawa, Optimal Finite- time Maxwell’s Demons in Langevin Systems , to appear in Physical Review Research
[19]

Knothe, Contributions to the theory of convex bodies , Michigan Mathematical J., 4 (1957), pp

H. Knothe, Contributions to the theory of convex bodies , Michigan Mathematical J., 4 (1957), pp. 39–52

1957
[20]

L´ eonard,From the Schr¨ odinger problem to the Monge–Kantorovich problem, J

C. L´ eonard,From the Schr¨ odinger problem to the Monge–Kantorovich problem, J. Funct. Anal., 262, no. 4 (2012), pp. 1879–1920

2012
[21]

L´ eonard,A survey of the Schr¨ odinger problem and some of its con- nections with optimal transport , Special Issue on Optimal Transport and Applications

C. L´ eonard,A survey of the Schr¨ odinger problem and some of its con- nections with optimal transport , Special Issue on Optimal Transport and Applications. Discrete Contin. Dyn. Syst., 34, no. 4 (2014), pp. 1533–1574. 27

2014
[22]

L´ evy, Th´ eorie de l’Addition des Variables Al´ eatoires, Gauthier- Villars, Paris, 1937

P. L´ evy, Th´ eorie de l’Addition des Variables Al´ eatoires, Gauthier- Villars, Paris, 1937

1937
[23]

Mikami, Monge’s problem with a quadratic cost by the zero-noise limit of h-path processes , Probab

T. Mikami, Monge’s problem with a quadratic cost by the zero-noise limit of h-path processes , Probab. Theory Related Fields, 129, no. 2 (2004), pp. 245–260

2004
[24]

T. Mikami, Variational characterization of the Knothe-Rosenblatt type rearrangements and its stochastic version, in Viscosity Solutions of Dif- ferential Equations and Related Topics, RIMS Kˆ okyˆ uroku, No. 1695, Kyoto University, Kyoto, 2010, pp. 35-55

2010
[25]

Mikami, A characterization of the Knothe–Rosenblatt processes by a convergence result , SIAM J

T. Mikami, A characterization of the Knothe–Rosenblatt processes by a convergence result , SIAM J. Control Optim. , 50, no. 4 (2012), pp. 1903–1920

2012
[26]

Mikami, Regularity of Schr¨ odinger’s functional equation in the weak topology and moment measures , J

T. Mikami, Regularity of Schr¨ odinger’s functional equation in the weak topology and moment measures , J. Math. Soc. Japan, 73 (2021), pp. 99–123

2021
[27]

Mikami, Stochastic Optimal Transportation: Stochastic Control with Fixed Marginals, SpringerBriefs in Mathematics, Springer , Sin- gapore, 2021

T. Mikami, Stochastic Optimal Transportation: Stochastic Control with Fixed Marginals, SpringerBriefs in Mathematics, Springer , Sin- gapore, 2021

2021
[28]

Mikami and M

T. Mikami and M. Thieullen, Duality theorem for stochastic optimal control problem, Stochastic Process. Appl., 113 (2006), pp. 1815–1835

2006
[29]

R. B. Nelsen, An Introduction to Copulas, 2nd ed., Springer , Heidel- berg, 2006

2006
[30]

Peyr´ e and M

G. Peyr´ e and M. Cuturi, Computational Optimal Transport: With Ap- plications to Data Science. Now Publishers, Boston (2019)

2019
[31]

Rachev and L

S.T. Rachev and L. R¨ uschendorf, Mass transportation problems, Vol. I: Theory, Vol. II: Application, Springer–Verlag, Berlin, 1998

1998
[32]

Rosenblatt, Remarks on a multivariate transformation , Annals of Mathematical Statistics, 23 (1952), pp

M. Rosenblatt, Remarks on a multivariate transformation , Annals of Mathematical Statistics, 23 (1952), pp. 470-472

1952
[33]

R¨ uschendorf and W

L. R¨ uschendorf and W. Thomsen, Note on the Schr¨ odinger equation and I-projections, Statist. Probab. Lett., 17 (1993), pp. 369–375. 28

1993
[34]

On the Reversal of the Laws of Nature

E. Schr¨ odinger, ¨Uber die Umkehrung der Naturgesetze , Sitz. Ber. der Preuss. Akad. Wissen., Berlin, Phys. Math., 8, no. 9 (1931), pp. 144–153. Translation: R. Chetrite, P. Muratore-Ginanneschi, and K. Schwieger, E. Schr¨ odinger’s 1931 paper “On the Reversal of the Laws of ...

1931
[35]

Schr¨ odinger,Th´ eorie relativiste de l’electron et l’interpr´ etation de la m´ ecanique quantique, Ann

E. Schr¨ odinger,Th´ eorie relativiste de l’electron et l’interpr´ etation de la m´ ecanique quantique, Ann. Inst. H. Poincar´ e, 2 (1932), pp. 269–310

1932
[36]

S. J. Sheu, Some estimates of the transition density of a nondegenerate diffusion Markov process, Ann. Probab., 19 (1991), pp. 538–561

1991
[37]

Talagrand, Transportation cost for Gaussian and other product mea- sures, Geom

M. Talagrand, Transportation cost for Gaussian and other product mea- sures, Geom. Funct. Anal., 6, no. 3 (1996), pp. 587–600

1996
[38]

Villani, Topics in optimal transportation, Providence, Amer

C. Villani, Topics in optimal transportation, Providence, Amer. Math. Soc., Providence, 2003

2003
[39]

J. C. Zambrini, Variational processes, in: Stochastic Processes in Clas- sical and Quantum Systems, Ascona 1985, Lecture Notes in Phys. 262, S. Albeverio, G. Casati, and D. Merlini, eds., Springer–Verlag, Heidel- berg, 1986, pp. 517–529

1985
[40]

J. C. Zambrini, Variational processes and stochastic versions of me- chanics, J. Math. Phys., 27 (1986), pp. 2307–2330. 29

1986

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

Beiglb¨ ock, G

M. Beiglb¨ ock, G. Pammer, and A. Posch,The Knothe–Rosenblatt dis- tance and its induced topology , arXiv:2312.16515

arXiv

[2] [2]

Bernstein, Sur les liaisons entre les grandeurs al´ etoires , Verh

S. Bernstein, Sur les liaisons entre les grandeurs al´ etoires , Verh. des intern. Mathematikerkongr. Zurich 1932, Band 1 (1932), pp. 288–309

work page 1932

[3] [3]

V. I. Bogachev, A. V. Kolesnikov, and K. V. Medvedev, On triangular transformations of measures (Russian), Dokl. Akad. Nauk, 396, no. 6 (2004), pp. 727–732

work page 2004

[4] [4]

V. I. Bogachev, A. V. Kolesnikov, and K. V. Medvedev, Triangular transformations of measures (Russian), Mat. Sb., 196, no. 3 (2005), pp. 3–30; translation in Sb. Math., 196, no. 3-4 (2005), pp. 309–335

work page 2005

[5] [5]

Carlier, F

G. Carlier, F. Galichon, and F. Santambrogio, From Knothe’s transport to Brenier’s map and a continuation method for optimal transport , SIAM J. Math. Anal., 41, no. 6 (2010), pp. 2554–2576

work page 2010

[6] [6]

Y. Chen, T. Georgiou, and M. Pavon, Entropic and displacement inter- polation: a computational approach using the Hilbert metric , SIAM J. Appl. Math., 76, no. 6 (2016), pp. 2375–2396

work page 2016

[7] [7]

Y. Chen, T. Georgiou, and M. Pavon, Stochastic control liaisons: Richard Sinkhorn meets Gaspard Monge on a Schr¨ odinger bridge. SIAM Review 63, no. 2, 249–313 (2021)

work page 2021

[8] [8]

Csiszar, I-divergence geometry of probability distributions and mini- mization problems, Ann

I. Csiszar, I-divergence geometry of probability distributions and mini- mization problems, Ann. Probab., 3 (1975), pp. 146–158

work page 1975