pith. sign in

arxiv: 2605.04246 · v1 · submitted 2026-05-05 · 🧮 math.OC · cs.LG· cs.RO· cs.SY· eess.SY

Globally Solving Unbalanced Optimal Transport and Density Control for Gaussian Distributions

Haruto Nakashima , Siddhartha Ganguly , Kenji Kashima This is my paper

Pith reviewed 2026-05-08 17:24 UTC · model grok-4.3

classification 🧮 math.OC cs.LGcs.ROcs.SYeess.SY
keywords unbalanced optimal transportunbalanced density controlGaussian measuressemidefinite programmingfinite dimensional reductionquadratic costKL divergence
0
0 comments X

The pith

Unbalanced optimal transport with Gaussian references reduces exactly to finite-dimensional optimization over masses, means and covariances.

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

The paper establishes that unbalanced optimal transport between prescribed Gaussian measures with quadratic costs and Kullback-Leibler penalties admits an exact reduction to a finite-dimensional problem. This reduction also applies to the dynamical extension called unbalanced density control on linear systems, where solutions can be restricted to Gaussian initial measures and affine-Gaussian policies. The resulting formulation is solved globally using semidefinite programming for fixed mass combined with a closed-form update for the mass. Sympathetic readers care because this bypasses the computational difficulties of infinite-dimensional variational problems for a practically relevant class of measures.

Core claim

The infinite-dimensional variational problem for unbalanced optimal transport admits an exact Gaussian reduction, yielding a finite-dimensional optimization over masses, means, and covariances, together with a closed-form expression for the optimal transported mass. For the unbalanced density control problem on discrete-time linear systems, any feasible solution can be replaced without loss of optimality by a Gaussian initial measure and an affine-Gaussian control policy, leading to an exact finite-dimensional reformulation that is solved via SDP-based optimization for fixed mass with a closed-form mass update.

What carries the argument

The exact Gaussian reduction that allows replacing any solution with a Gaussian measure and affine-Gaussian policy, enabling SDP reformulation.

If this is right

  • Global optimality for Gaussian unbalanced optimal transport is achieved through the finite-dimensional SDP.
  • A closed-form expression exists for the optimal transported mass.
  • Existence of optimal solutions is guaranteed for the problems considered.
  • A sufficient condition identifies when the affine-Gaussian policy is deterministic.

Where Pith is reading between the lines

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

  • The reduction suggests that similar exact solvability holds whenever the cost structure preserves Gaussianity under linear dynamics.
  • Computationally, this turns an intractable variational problem into a tractable SDP that scales with dimension rather than measure space.

Load-bearing premise

The reference measures must be Gaussian and the dynamics discrete-time linear with quadratic control costs and KL penalties on the marginals.

What would settle it

An instance with non-Gaussian reference measures where the finite-dimensional Gaussian solution fails to achieve the true optimal transport cost.

Figures

Figures reproduced from arXiv: 2605.04246 by Haruto Nakashima, Kenji Kashima, Siddhartha Ganguly.

Figure 1
Figure 1. Figure 1: The optimal marginal distribution in Case 1 (top) and Case 2 (bottom), where 𝛾 = 0.2 and 𝛾 = 30, respectively. The dotted orange line indicates the optimal marginal distribution, and the blue line specifies the given reference measure view at source ↗
Figure 2
Figure 2. Figure 2: Optimal transport plans for the UOT problem with 𝛾 = 0.2 (left), 𝛾 = 30 (middle), and for standard OT (right). OT coupling. Consistently, view at source ↗
Figure 3
Figure 3. Figure 3: The reference measures 𝛼 and 𝛽 with the optimal state measure 𝜋𝑘 for UDC problem when 𝛾 = 3.0 (left) and 𝛾 = 10 (right), computed by Algorithm 2. As 𝛾 increases, the optimized initial and terminal measures move closer to the prescribed reference measures. the transport cost. The first term in (3) favors transport over short distances and penalizes transport over longer distances; accordingly, the cost is s… view at source ↗
Figure 4
Figure 4. Figure 4: (Top) The optimal transport plans for the entropic UOT prob￾lem. For 𝛾 = 30, the left-hand sub figure depicts the result when 𝜎 = 0.1, and the right-hand subfigure when 𝜎 = 0.5. (Right) 200 samples of the optimal state process for (49) and (50). For fixed 𝛾 = 1.0, the left-hand subfigure depicts the result when 𝜀 = 0.02, and the right-hand figure when 𝜀 = 0.4. the additional entropy regularization term, wh… view at source ↗
read the original abstract

In this article, we study unbalanced optimal transport (UOT) and establish a control-theoretic dynamical extension, which we call the unbalanced density control (UDC), for a class of Gaussian reference measures. In the static setting, we consider UOT with quadratic transport cost and Kullback--Leibler penalties on the marginals relative to prescribed Gaussian measures. We show that the infinite-dimensional variational problem admits an exact Gaussian reduction, yielding a finite-dimensional optimization over masses, means, and covariances, together with a closed-form expression for the optimal transported mass. We then formulate UDC for discrete-time linear systems, where the initial and terminal state measures are imposed softly through KL penalties and the intermediate evolution is governed by controlled linear dynamics with quadratic control cost. For this problem, we prove that any feasible solution can be replaced, without loss of optimality, by a Gaussian initial measure and an affine-Gaussian control policy. This leads to an exact finite-dimensional reformulation and, after a standard covariance-steering lifting, to an SDP-based optimization for fixed mass, again coupled with a closed-form mass update. We further establish existence of optimal solutions and identify a sufficient condition under which the affine-Gaussian UDC policy is deterministic. These results provide globally optimal solution methods for both Gaussian UOT and Gaussian UDC. Finally, we illustrate our results with several numerical examples.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 2 minor

Summary. The manuscript studies unbalanced optimal transport (UOT) with quadratic transport cost and KL penalties relative to Gaussian reference measures, claiming an exact reduction of the infinite-dimensional variational problem to a finite-dimensional optimization over masses, means, and covariances together with a closed-form expression for the optimal transported mass. It then extends this to unbalanced density control (UDC) on discrete-time linear systems, where initial and terminal measures are penalized softly via KL divergences and the dynamics are controlled with quadratic cost; the central claim is that any feasible solution can be replaced without loss of optimality by a Gaussian initial measure and an affine-Gaussian policy. This yields an exact finite-dimensional reformulation that, after covariance-steering lifting, becomes an SDP for fixed mass coupled with a closed-form mass update. The authors also prove existence of optimal solutions and give a sufficient condition for the optimal policy to be deterministic, with numerical illustrations.

Significance. If the exactness of the Gaussian reduction and the replacement lemma hold, the work would be significant: it supplies globally optimal, computationally tractable methods (via SDP solvers plus closed-form updates) for infinite-dimensional UOT and UDC problems in the Gaussian setting. The combination of quadratic costs, linear dynamics, and KL penalties to Gaussians is exploited to obtain parameter-free finite-dimensional equivalents, which is a strong technical achievement when the derivations are rigorous.

major comments (2)
  1. [Gaussian replacement lemma (proof section)] The Gaussian replacement lemma (abstract and the section containing its proof): the claim that every feasible initial measure and control policy can be replaced, without loss of optimality, by a Gaussian initial measure and an affine-Gaussian policy is load-bearing for the entire finite-dimensional reformulation and SDP reduction. The argument presumably uses the quadratic cost, linear dynamics, and the fact that KL divergences to Gaussians are preserved or improved under affine maps, but it is not obvious that the KL term for an arbitrary non-Gaussian measure can always be matched or strictly decreased while keeping the identical control cost and terminal marginal. If the replacement only yields a lower bound for some non-Gaussian cases, the SDP would solve a relaxation rather than the original problem.
  2. [SDP reformulation paragraph] Covariance-steering lifting step (the paragraph after the replacement lemma): after restricting to Gaussian measures and affine policies, the lifting to an SDP for fixed mass is presented as standard, yet the precise semidefinite constraints on the lifted covariance variables and the coupling with the closed-form mass update are not cross-referenced to an equation number. Without an explicit statement of the lifted SDP (e.g., the matrix inequalities involving the lifted second-moment variables), it is difficult to verify that the lifting is exact and that the subsequent mass update recovers the global optimum.
minor comments (2)
  1. [Problem formulation] Notation for the unbalanced penalties: the KL terms are written with respect to prescribed Gaussian measures, but the precise scaling constants multiplying the KL divergences (if any) are not displayed in the abstract formulation; a single displayed equation collecting all terms of the objective would improve readability.
  2. [Numerical examples] Numerical examples: the figures show trajectories or marginals but do not report the achieved objective values or compare against a non-Gaussian baseline solver; adding such quantitative verification would help confirm that the computed solutions are indeed globally optimal.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive review. The comments raise important points about the rigor of the Gaussian replacement lemma and the explicitness of the SDP formulation. We address each below and indicate planned revisions.

read point-by-point responses

  1. Referee: [Gaussian replacement lemma (proof section)] The Gaussian replacement lemma (abstract and the section containing its proof): the claim that every feasible initial measure and control policy can be replaced, without loss of optimality, by a Gaussian initial measure and an affine-Gaussian policy is load-bearing for the entire finite-dimensional reformulation and SDP reduction. The argument presumably uses the quadratic cost, linear dynamics, and the fact that KL divergences to Gaussians are preserved or improved under affine maps, but it is not obvious that the KL term for an arbitrary non-Gaussian measure can always be matched or strictly decreased while keeping the identical control cost and terminal marginal. If the replacement only yields a lower bound for some non-Gaussian cases, the SDP would solve a relaxation rather than the original problem.

    Authors: We appreciate the referee's scrutiny of this foundational result. The lemma establishes exact equivalence rather than a relaxation. For any feasible initial measure μ, let ν be the Gaussian sharing the same mean and covariance. The quadratic control cost and linear dynamics imply that the expected cost depends only on first- and second-order moments; these moments are preserved under an appropriately chosen affine-Gaussian policy, so the control cost remains identical. For the KL penalty, the reference measure is Gaussian, so KL(μ || N_ref) = -H(μ) + quadratic moment term + constant. By the maximum-entropy property, H(μ) ≤ H(ν) with equality iff μ is Gaussian; consequently KL(μ || N_ref) ≥ KL(ν || N_ref). The terminal KL term is handled analogously by matching moments at the terminal time. Thus the objective value attained by the Gaussian replacement is at most that of the original pair (μ, policy), proving that the global optimum lies within the Gaussian-affine class. We will expand the proof section with these explicit entropy and moment arguments to make the reasoning fully transparent. revision: partial

  2. Referee: [SDP reformulation paragraph] Covariance-steering lifting step (the paragraph after the replacement lemma): after restricting to Gaussian measures and affine policies, the lifting to an SDP for fixed mass is presented as standard, yet the precise semidefinite constraints on the lifted covariance variables and the coupling with the closed-form mass update are not cross-referenced to an equation number. Without an explicit statement of the lifted SDP (e.g., the matrix inequalities involving the lifted second-moment variables), it is difficult to verify that the lifting is exact and that the subsequent mass update recovers the global optimum.

    Authors: We agree that the SDP lifting paragraph would benefit from greater explicitness. In the revised version we will insert numbered equations that state the lifted SDP explicitly: the decision variables include the lifted second-moment matrix Z together with the mean vectors; the constraints comprise the linear matrix inequality Z ≽ [m; m^T, Σ] (or its equivalent Schur form), the discrete-time Lyapunov-type equalities propagating the second moments under the affine policy, and the positive-semidefiniteness conditions on the covariance blocks. We will also add a cross-reference showing how the optimal mass is recovered in closed form once the SDP is solved for each fixed mass value, confirming that the procedure yields the global optimum of the finite-dimensional problem. revision: yes

Circularity Check

0 steps flagged

No circularity: derivations rest on explicit proofs of Gaussian replacement and SDP lifting rather than self-definition or fitted inputs

full rationale

The paper's central steps are (1) a claimed exact Gaussian reduction for the static UOT variational problem and (2) a replacement lemma asserting that any feasible (possibly non-Gaussian) initial measure and control policy for the discrete-time linear UDC problem can be replaced by a Gaussian measure plus affine-Gaussian policy without increasing the objective. These are presented as theorems proved from the quadratic cost, linear dynamics, and KL penalties to Gaussian references. No parameter is fitted to data and then relabeled a prediction; no ansatz is smuggled via self-citation; the replacement lemma is not defined in terms of the finite-dimensional SDP it enables. The subsequent covariance-steering lift to an SDP for fixed mass plus closed-form mass update follows standard techniques once the Gaussian restriction is justified. The derivation chain is therefore self-contained against the stated assumptions and does not reduce to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The work relies on standard closed-form properties of Gaussians under affine maps and KL divergence, plus existence results for linear-quadratic control problems. No new free parameters or invented entities are introduced in the abstract.

axioms (2)
  • standard math Gaussian distributions remain Gaussian under affine transformations and admit closed-form expressions for KL divergence and quadratic transport costs.
    Invoked to obtain the exact finite-dimensional reduction.
  • standard math The covariance-steering problem for linear systems with quadratic cost admits an SDP formulation.
    Used in the dynamic reformulation step.

pith-pipeline@v0.9.0 · 5561 in / 1440 out tokens · 48670 ms · 2026-05-08T17:24:05.759571+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

112 extracted references · 112 canonical work pages

  1. [1]

    Positivity of 2 2 block matrices of operators , volume=

    Moslehian, Mohammad Sal and Kian, Mohsen and Xu, Qingxiang , year=. Positivity of 2 2 block matrices of operators , volume=. Banach Journal of Mathematical Analysis , publisher=

    work page

  2. [2]

    Papadakis , title =

    N. Papadakis , title =. 2015 , NOTE =

    work page 2015

  3. [3]

    Gazdieva and P

    M. Gazdieva and P. Mokrov and L. Rout and A. Korotin and A. Kravchenko and A. Filippov and E. Burnaev , title =. Journal of Optimization Theory and Applications , volume =. 2025 , publisher =

    work page 2025

  4. [4]

    Chen and T

    Y. Chen and T. T. Georgiou and M. Pavon , journal=. On the relation between optimal transport and. 2016 , publisher=

    work page 2016

  5. [5]

    Bon and G

    D. Bon and G. Pai and G. Bellaard and O. Mula and R. Duits , title =. SIAM Journal on Imaging Sciences , volume =. 2025 , publisher =

    work page 2025

  6. [6]

    G. Peyr. Optimal and diffusion transports in machine learning , year =

    work page

  7. [7]

    J. D. Benamou and Y. Brenier , title =. Numerische Mathematik , volume =. 2000 , publisher =

    work page 2000

  8. [8]

    Monge , title =

    G. Monge , title =. M. 1781 , NOTE =

    work page

  9. [9]

    Gelbrich , title =

    M. Gelbrich , title =. Mathematische Nachrichten , volume =. 1990 , NOTE =

    work page 1990

  10. [10]

    C. R. Givens and R. M. Shortt , title =. Michigan Math. J , volume =. 1984 , NOTE =

    work page 1984

  11. [11]

    Cover and Joy A

    Thomas M. Cover and Joy A. Thomas , title =. Wiley-Interscience , year =

    work page

  12. [12]

    G. Peyr. Computational. 2019 , publisher=

    work page 2019

  13. [13]

    T. S. Unbalanced optimal transport, from theory to numerics , journal =. 2023 , NOTE =

    work page 2023

  14. [14]

    Chizat and G

    L. Chizat and G. Peyré and B. Schmitzer and F.-X. Vialard , keywords =. Unbalanced optimal transport: Dynamic and. Journal of Functional Analysis , volume =. 2018 , issn =

    work page 2018

  15. [15]

    Pham and K

    K. Pham and K. Le and N. Ho and T. Pham and H. Bui , title =. Proceedings of the 37th International Conference on Machine Learning , series =. 2020 , NOTE =

    work page 2020

  16. [16]

    2026 , eprint=

    Covariance Steering of Discrete-Time Markov Jump Linear Systems with Multiplicative Noise , author=. 2026 , eprint=

    work page 2026

  17. [17]

    Zavlanos and and Karl H

    Zifan Wang and Yi Shen and Michael M. Zavlanos and and Karl H. Johansson , title=. Advances in Neural Information Processing Systems , year=

    work page

  18. [18]

    Data-driven

    Haruto Nakashima and Siddhartha Ganguly and Kenji Kashima , booktitle=. Data-driven. 2025 , volume=

    work page 2025

  19. [19]

    Foundations of Computational Mathematics , year=

    Gabriel Peyré and Bernhard Schmitzer and François-Xavier Vialard , title=. Foundations of Computational Mathematics , year=

    work page

  20. [20]

    IEEE Control Systems Letters , year=

    Kohei Morimoto and Kenji Kashima , title=. IEEE Control Systems Letters , year=

    work page

  21. [21]

    IEEE Transactions on Automatic Control , year=

    Fengjiao Liu and George Rapakoulias and Panagiotis Tsiotras , title=. IEEE Transactions on Automatic Control , year=

    work page

  22. [22]

    Georgiou and Michele Pavon , journal=

    Yongxin Chen and Tryphon T. Georgiou and Michele Pavon , journal=. Stochastic control liaisons: Richard. 2021 , publisher=

    work page 2021

  23. [23]

    T. M. Cover and J. A. Thomas , title =. 2006 , NOTE =

    work page 2006

  24. [24]

    Gallier , title =

    J. Gallier , title =. 2011 , NOTE =

    work page 2011

  25. [25]

    Ambrosio and G

    L. Ambrosio and G. Buttazzo , title =. Annali di Matematica Pura ed Applicata , volume =. 1988 , publisher =

    work page 1988

  26. [26]

    Morimoto and K

    K. Morimoto and K. Kashima , title =. Proceedings of Machine Learning Research , volume =. 2025 , NOTE =

    work page 2025

  27. [27]

    McCan , title =

    Robert J. McCan , title =. Advances in Mathematics , volume=. 1997 , NOTE =

    work page 1997

  28. [28]

    I. M. Balci and E. Bakolas , title=. 2022 , eprint=

    work page 2022

  29. [29]

    Hicham and B

    J. Hicham and B. Muzellec and G. Peyré and M. Cuturi , title =. Advances in neural information processing systems 33 , year =

    work page

  30. [30]

    Santambrogio , year=

    F. Santambrogio , year=. A

    work page

  31. [31]

    Optimal Entropy-Transport problems and a new

    Liero, Matthias and Mielke, Alexander and Savaré, Giuseppe , year=. Optimal Entropy-Transport problems and a new. Inventiones mathematicae , publisher=

    work page

  32. [32]

    Ito and K

    K. Ito and K. Kashima , title =. IEEE Transactions on Automatic Control , year =

    work page

  33. [33]

    Villani , TITLE =

    C. Villani , TITLE =. 2009 , PAGES =

    work page 2009

  34. [34]

    Albert , journal =

    A. Albert , journal =. Conditions for Positive and Nonnegative Definiteness in Terms of Pseudoinverses , volume =. 1969 , NOTE =

    work page 1969

  35. [35]

    Direct methods in the calculus of variations , volume =

    Garguichevich Graciela and Gariboldi Claudia and Marangunic Pedro and Pallara Diego , year =. Direct methods in the calculus of variations , volume =. MAT Serie A , doi =

    work page

  36. [36]

    Gallier , year=

    J. Gallier , year=. Geometric

    work page

  37. [37]

    Annali di Matematica Pura ed Applicata , volume=

    Weak lower semicontinuous envelope of functionals defined on a space of measures , author=. Annali di Matematica Pura ed Applicata , volume=. 1988 , publisher=

    work page 1988

  38. [38]

    Mémoire sur la théorie des déblais et des remblais , author=

    work page

  39. [39]

    On the translocation of masses , author=. Dokl. Akad. Nauk. USSR (NS) , volume=. 1942 , NOTE =

    work page 1942

  40. [40]

    Villani , title =

    C. Villani , title =. 2009 , isbn =

    work page 2009

  41. [41]

    Numerische Mathematik , volume=

    A computational fluid mechanics solution to the Monge-Kantorovich mass transfer problem , author=. Numerische Mathematik , volume=. 2000 , publisher=

    work page 2000

  42. [42]

    K. F. Caluya and A. Halder , journal=. Wasserstein proximal algorithms for the Schr. 2021 , publisher=

    work page 2021

  43. [43]

    P. M. Esfahani and D. Kuhn , journal=. Data-driven distributionally robust optimization using the. 2018 , publisher=

    work page 2018

  44. [44]

    Halder and E

    A. Halder and E. D. B. Wendel , booktitle=. Finite horizon linear quadratic Gaussian density regulator with. 2016 , volume=

    work page 2016

  45. [45]

    2025 , eprint=

    Fixed Horizon Linear Quadratic Covariance Steering in Continuous Time with Hilbert-Schmidt Terminal Cost , author=. 2025 , eprint=

    work page 2025

  46. [46]

    Pilipovsky and P

    J. Pilipovsky and P. Tsiotras , booktitle=. Distributionally Robust Density Control with. 2024 , volume=

    work page 2024

  47. [47]

    Journal of Optimization Theory and Applications , volume=

    An optimal transport perspective on unpaired image super-resolution , author=. Journal of Optimization Theory and Applications , volume=. 2025 , publisher=

    work page 2025

  48. [48]

    Bon and G

    D. Bon and G. Pai and G. Bellaard and O. Mula and R. Duits , journal=. Optimal transport on the. 2025 , publisher=

    work page 2025

  49. [49]

    Chen and T

    Y. Chen and T. T. Georgiou and M. Pavon , journal=. Optimal steering of a linear stochastic system to a final probability distribution, Part. 2015 , NOTE =

    work page 2015

  50. [50]

    Chen and T

    Y. Chen and T. T. Georgiou and M. Pavon , journal=. Optimal Steering of a Linear Stochastic System to a Final Probability Distribution—Part. 2018 , volume=

    work page 2018

  51. [51]

    Eldesoukey and O

    A. Eldesoukey and O. M. Miangolarra and T. T. Georgiou , journal=. An Excursion Onto. 2024 , volume=

    work page 2024

  52. [52]

    Chen and T

    Y. Chen and T. T. Georgiou and M. Pavon , title =. SIAM Review , volume =. 2025 , NOTE =

    work page 2025

  53. [53]

    SIAM Journal on Control and Optimization , volume=

    The most likely evolution of diffusing and vanishing particles: Schrodinger bridges with unbalanced marginals , author=. SIAM Journal on Control and Optimization , volume=. 2022 , publisher=

    work page 2022

  54. [54]

    European Journal of Applied Mathematics , volume=

    Interpolation of matrices and matrix-valued densities: The unbalanced case , author=. European Journal of Applied Mathematics , volume=. 2019 , publisher=

    work page 2019

  55. [55]

    International Journal of Control , volume=

    Covariance control theory , author=. International Journal of Control , volume=. 1987 , publisher=

    work page 1987

  56. [56]

    24th IEEE Conference on Decision and Control , pages=

    Covariance control discrete systems , author=. 24th IEEE Conference on Decision and Control , pages=. 1985 , NOTE =

    work page 1985

  57. [57]

    IEEE Transactions on Automatic Control , volume=

    An improved covariance assignment theory for discrete systems , author=. IEEE Transactions on Automatic Control , volume=. 2002 , NOTE =

    work page 2002

  58. [58]

    Automatica , volume=

    Minimum-energy covariance controllers , author=. Automatica , volume=. 1997 , publisher=

    work page 1997

  59. [59]

    Automatica , volume=

    Finite-horizon covariance control for discrete-time stochastic linear systems subject to input constraints , author=. Automatica , volume=. 2018 , publisher=

    work page 2018

  60. [60]

    SIAM Journal on Control and Optimization , volume=

    Convex optimization for finite-horizon robust covariance control of linear stochastic systems , author=. SIAM Journal on Control and Optimization , volume=. 2021 , publisher=

    work page 2021

  61. [61]

    IEEE 58th Conference on Decision and Control (CDC) , pages=

    Input hard constrained optimal covariance steering , author=. IEEE 58th Conference on Decision and Control (CDC) , pages=. 2019 , NOTE =

    work page 2019

  62. [62]

    IEEE Transactions on Automatic Control , volume=

    Optimal covariance steering for discrete-time linear stochastic systems , author=. IEEE Transactions on Automatic Control , volume=. 2024 , NOTE =

    work page 2024

  63. [63]

    I. M. Balci and E. Bakolas , journal=. Exact. 2022 , NOTE =

    work page 2022

  64. [64]

    Sial and A

    T. Sial and A. Halder , year=. Fixed Horizon Linear Quadratic Covariance Steering in Continuous Time with. 2510.21944 , archivePrefix=

    work page arXiv

  65. [65]

    Chapel and R

    L. Chapel and R. Flamary and H. Wu and C. F\'. Unbalanced Optimal Transport through Non-negative Penalized Linear Regression , NOTE =. Advances in Neural Information Processing Systems , editor =

    work page

  66. [66]

    Chance-Constrained Covariance Steering for Discrete-Time

    Shrivastava, Shaurya and Oguri, Kenshiro , journal=. Chance-Constrained Covariance Steering for Discrete-Time. 2025 , NOTE =

    work page 2025

  67. [67]

    arXiv preprint arXiv:2406.14740 , year=

    Reachability and controllability analysis of the state covariance for linear stochastic systems , author=. arXiv preprint arXiv:2406.14740 , year=

    work page arXiv

  68. [68]

    62nd IEEE Conference on Decision and Control (CDC) , pages=

    Data-driven covariance steering control design , author=. 62nd IEEE Conference on Decision and Control (CDC) , pages=. 2023 , NOTE =

    work page 2023

  69. [69]

    Ono and B

    M. Ono and B. C. Williams , booktitle=. Iterative Risk Allocation: A new approach to robust Model Predictive Control with a joint chance constraint , year=

    work page

  70. [70]

    J. W. Knaup and P. Tsiotras , booktitle=. Computationally Efficient Covariance Steering for Systems Subject to Parametric Disturbances and Chance Constraints , year=

    work page

  71. [71]

    Nakashima and S

    H. Nakashima and S. Ganguly and K. Morimoto and K. Kashima , booktitle=. Formation Shape Control using the. 2025 , volume=

    work page 2025

  72. [72]

    Okamoto and M

    K. Okamoto and M. Goldshtein and P. Tsiotras , journal=. Optimal Covariance Control for Stochastic Systems Under Chance Constraints , year=

    work page

  73. [73]

    Ye and Z

    L. Ye and Z. Zhao and F. Liu , journal =. Stochastic. 2024 , NOTE =

    work page 2024

  74. [74]

    Systems & Control Letters , volume =

    Constrained quadratic control for. Systems & Control Letters , volume =. 2025 , author =

    work page 2025

  75. [75]

    O. L. V. Costa and W. L. de Paulo , journal=. Indefinite quadratic with linear costs optimal control of. 2007 , publisher=

    work page 2007

  76. [76]

    Fast Switching Detector-Based

    Graciani Rodrigues, CC and Todorov, Marcos G and Fragoso, Marcelo D , journal=. Fast Switching Detector-Based. 2021 , publisher=

    work page 2021

  77. [77]

    Schuurmans and P

    M. Schuurmans and P. Patrinos , journal=. A general framework for learning-based distributionally robust. 2023 , NOTE =

    work page 2023

  78. [78]

    Chitraganti and S

    S. Chitraganti and S. Aberkane and C. Aubrun and G. Valencia-Palomo and V. Dragan , NOTE =. On control of discrete-time state-dependent jump linear systems with probabilistic constraints: A receding horizon approach , journal =. 2014 , issn =

    work page 2014

  79. [79]

    Barbieri and O

    F. Barbieri and O. L. V. Costa , journal=. Optimal control with constrained total variance for. 2018 , publisher=

    work page 2018

  80. [80]

    O. L. V. Costa and R. P. Marques and M. D. Fragoso , year=. Discrete-time

    work page

Showing first 80 references.