Pith. sign in

REVIEW 3 major objections 4 minor 44 references

Long-time behavior and turnpike properties of linear-quadratic graphon mean field control problems

T0 review · 3 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read For linear-quadratic graphon mean field control, the finite-horizon optimal state-control pair converges exponentially fast to the optimal pair of the associated ergodic problem, uniformly in the horizon length and allowing different initia

desk verdict A first turnpike theorem for graphon mean field control, but the headline result rides on an assumed convergence that Section 6 does not actually prove. read the letter →

arxiv 2607.18000 v1 pith:USMMVRXD submitted 2026-07-20 math.OC math.PR

Erhan Bayraktar , Zhongyuan Cao , Jiamin Jian This is my paper
classification math.OCmath.PR MSC 93E2049N1093E1549N9034H05
keywords graphonmeanfieldcontrolturnpikepropertylinear-quadraticRiccatiequationsheterogeneousinteractionsergodiclong-timebehaviorstabilizability
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper studies a social-planner optimal control problem for a continuum of agents with heterogeneous pairwise interactions encoded by a graphon kernel. For a linear-quadratic cost, it proves that as the time horizon grows, the optimal trajectory and control stay exponentially close to those of a stationary ergodic problem over most of the interval, with an explicit two-sided exponential bound in t and T−t. This turnpike phenomenon is derived from exponential convergence of the associated Riccati equations to algebraic steady-state equations on Hilbert spaces. The authors also show the time-averaged finite-horizon value converges to the ergodic value. A sympathetic reader cares because this reduces difficult long-horizon heterogeneous mean-field planning to a tractable stationary problem.

What carries the argument

The load-bearing objects are two Riccati towers. For the finite-horizon problem: a standard matrix Riccati ODE for P_T, another for Π̄_T, an abstract Riccati equation on L²(I×I) for the kernel Λ_T, and linear equations for p_T and κ_T. For the ergodic problem: algebraic Riccati equations for P and Π̄, an abstract algebraic equation for Λ, and a linear equation for p. The turnpike result is obtained by proving exponential estimates comparing the time-dependent solutions to their stationary limits, using stabilizability and the exponentially stable operator L built from the ergodic feedback gains, then transferring these estimates to the state-control gap via variation of constants, Lyapunov e

What would settle it

For a stabilizable, positive example with a finite-rank graphon, numerically solve the finite-horizon abstract Riccati ODE (8) and the algebraic equation (24) and check whether Λ_T(t) converges to Λ in L² as T→∞ for a fixed interior t. A single such instance where the finite-horizon solution stays bounded away from the algebraic solution as T grows would show that the stated exponential turnpike bound cannot hold in its present form.

Watch

Extended reading notes

Core claim

The central claim, Theorem 5.1, is that for the linear-quadratic graphon mean field control problem, the optimal pair (state law and control) of the finite-horizon problem converges exponentially fast to the optimal pair of the associated ergodic problem, uniformly in the horizon T and allowing possibly different initial states. The error is bounded by K(e^{-λt}+e^{-λ(T−t)}) in the label-averaged L² norm, with K and λ independent of T, so that away from the boundary layers at 0 and T the two solutions coincide to within an exponentially small error. The proof goes through exponential convergence of the finite-horizon Riccati system to the ergodic algebraic system, and also yields convergence

Load-bearing premise

The argument leans on Assumption C's second half: the finite-horizon Riccati kernel Λ_T(t) must converge to the ergodic limit Λ as T grows; if that convergence fails, the exponential Riccati estimates and Theorem 5.1 lose their support.

Editorial extensions

If this is right

  • Long-horizon planning in heterogeneous graphon mean field control can be split into solving one ergodic algebraic system and patching two boundary layers of width O(1/λ).
  • The optimal feedback gains of the finite-horizon problem converge exponentially, as the horizon recedes, to the ergodic feedback gains, so the stationary feedback law is a near-optimal policy for long horizons.
  • The time-averaged value of the finite-horizon problem converges to the initial-state-independent ergodic value.
  • When the graphon is bounded, the turnpike estimate holds pointwise in the agent label, not only in the aggregate L² sense.
  • Under stabilizability and the paper's positivity conditions, the ergodic GMFC problem is uniquely solvable and its optimal value is independent of the initial state.

Reading between the lines

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

  • The exponential rate λ controls the boundary-layer thickness; estimating or bounding λ could tell practitioners how long a horizon must be before the stationary solution becomes a reliable approximation.
  • The same Riccati-comparison strategy may extend to graphon mean field games or to weakly nonlinear graphon control via local linearization, though the algebraic equations would lose their quadratic structure.
  • Because the turnpike bound allows different initial states, it suggests an attracting or memoryless property of the ergodic optimum: initial conditions affect the solution only through the two boundary layers.
  • Making the sufficient condition in Section 6 checkable for concrete graphon classes, such as low-rank or smooth kernels, would turn the convergence assumption into a testable hypothesis.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper studies linear-quadratic graphon mean field control (GMFC) problems over finite and infinite horizons. For the finite-horizon problem it derives a system of generalized Riccati differential equations and, for the ergodic problem, an associated algebraic system. Under Assumptions A, B and C it proves exponential convergence of the finite-horizon Riccati kernels to their ergodic limits (Prop. 4.2), establishes an exponential turnpike property for the optimal state‐control pair with possibly different initial states (Theorem 5.1), and shows convergence of the time-averaged finite-horizon value to the ergodic value (Lemma 5.3). Section 6 proposes an operator matrix inequality, Assumption D, as a sufficient condition for the unique solvability of the algebraic equations and for the qualitative convergence required by Assumption C. The main structural weakness is that Assumption C already contains the qualitative L2 convergence of the finite-horizon Riccati kernel to its algebraic limit; Proposition 4.2 then upgrades this qualitative input to an exponential rate, while Section 6's attempt to derive C from D contains an invalid Hilbert-space limit passage.

Significance. If the main theorem were established from transparent hypotheses, this would be the first turnpike result for graphon mean field control problems and would be a useful contribution to the growing literature on long-time behavior of LQ mean field control. The paper is careful in setting up the graphon-indexed Riccati systems, in deriving the fundamental relations for both finite-horizon and ergodic costs, and in organizing the coefficient convergence estimates. The verification theorems and the Riccati flow are standard in spirit and the algebra is carried out in detail. However, the advertised result is not obtained from stabilizability and positivity alone: the central theorem is conditional on Assumption C, whose second half already asserts the qualitative convergence that Proposition 4.2 later claims to strengthen. Moreover, the only proposed route to verify Assumption C, Proposition 6.2, has a functional-analytic gap. These issues are load-bearing for the paper's central claim, so the paper cannot be accepted in its present form.

major comments (3)
  1. [§3.3, Assumption C; §4, Proposition 4.2; §5.1, Theorem 5.1] Assumption C explicitly includes, for each fixed t≥0, the L2 convergence Λ_T(t)→Λ as T→∞. In the proof of Proposition 4.2 this convergence is invoked directly: 'By Assumption C ... lim ∥ΔΛ(0)∥=0' (proof around Eq. (37)). The argument then bootstraps this qualitative convergence to the exponential estimate (37). Therefore Theorem 5.1 does not establish a turnpike property from the stabilizability/positivity assumptions A and B; it is conditional on an assumption that already contains a version of the long-time convergence of the Riccati kernel. The paper should state this dependency prominently, or replace Assumption C by a hypothesis that does not itself encode the qualitative convergence being derived.
  2. [§6, Proposition 6.2] The proof asserts that from monotonicity and the uniform quadratic-form bound ⟨ξ,eΛ(t)ξ⟩≤K∥ξ∥² it follows that eΛ(t)=Λ_T(T−t) converges in L2(I×I;R^{d×d}) to some eΛ_∞. This is not valid: monotone uniform boundedness of bounded self-adjoint operators yields strong operator convergence to a bounded self-adjoint operator, not convergence in Hilbert–Schmidt norm. A uniform bound on the Hilbert–Schmidt norms ∥eΛ(t)∥_{L2(I×I)} is needed, and none is supplied. The subsequent 'passing to the limit in the finite-horizon Riccati equation (8)' is therefore not justified. Since Proposition 6.2 is the only mechanism proposed to discharge the convergence half of Assumption C, the claimed sufficiency of Assumptions A, B, D for Theorem 5.1 is not established.
  3. [§6, Proposition 6.2 (continuation)] Even if the L2 convergence of eΛ(t) were established, the passage to the limit in equation (8) requires convergence of the nonlinear terms F(s,Λ_T(s)) in the appropriate L2 sense. The proof does not provide such a convergence argument, e.g., via a uniform Lipschitz estimate in L2 and a dominated-convergence argument. The uniqueness part of (24) is also only sketched, since it is inferred from the limiting procedure rather than from a direct monotonicity or coercivity property of the algebraic equation. This reinforces that Section 6 is not yet a rigorous sufficient condition.
minor comments (4)
  1. [Abstract and §1.1] The abstract says 'under a stabilizability condition and appropriate positivity assumptions' but does not mention that the turnpike theorem uses Assumption C, which contains a convergence property. The introduction similarly says the main theorem is proved 'under a suitable stabilizability condition and appropriate positivity assumptions'; this should be qualified to avoid overstating the hypotheses.
  2. [Remark 2.1] The comparison with [19] uses Dirac delta graphons that are distributions rather than L2 kernels; the paper correctly notes this, but the wording 'graphons in our formulation are distributions' could be clarified as being a formal identification only, since the paper's framework itself requires L2 graphons.
  3. [§4, Proposition 4.2] The notation ΔarΘ^*(t) is used before being defined (it is defined as arΘ^* - arΘ^*_T(t) in the text after the equation for ΔΛ). Please define all Δ-quantities in one place before the proof.
  4. [Proof of Lemma 5.3] The estimate (45) uses boundedness of ∥P_T(t)∥, ∥R(P_T(t))^{-1}∥ and ∥p_T(t)∥_{L2} uniformly in t and T. This is true under the preceding estimates, but it would be helpful to state it explicitly before the final convergence step, since the argument depends on it.

Circularity Check

1 steps flagged · score 7.0 of 10

Assumption C already contains the qualitative Riccati convergence that Proposition 4.2 'proves' exponentially, so the central turnpike theorem is conditional on its own conclusion.

  1. self definitional [Assumption C (Section 3.3) and Proposition 4.2 (Section 4, proof of (37))]
    "Moreover, for each fixed t≥0, the solution Λ_T(t) of (8) converges to the solution Λ of (24) in L²(I×I;R^{d×d}) as T→∞. ... By AssumptionC, for each fixed t≥0, Λ_T(t) converges to Λ in L²(I×I;R^{d×d}) as T→∞. Specifically, lim_{T→∞} ∥∆Λ(0)∥_{L²}=0."

    Proposition 4.2 is presented as deriving the exponential convergence estimate (37), but its proof begins by invoking the qualitative L² convergence already imposed as the second half of Assumption C. Thus the convergence of the finite-horizon Riccati kernel to its ergodic limit—the key analytic input to Theorem 5.1—is an assumed input rather than a consequence of stabilizability (Assumption B) and positivity (Assumption A). Only the exponential rate is genuinely derived, by bootstrapping around an assumed limit. Since Theorem 5.1 is stated under Assumption C, the main turnpike claim is conditional on its own qualitative conclusion.

full rationale

The circularity burden is concentrated in Assumption C. Its second clause asserts the qualitative L² convergence Λ_T(t)→Λ that Proposition 4.2 then 'proves' with an exponential rate; the proof explicitly uses Assumption C to obtain the input needed for the rate bootstrap. The exponential estimates (37)–(40) feed directly into Theorem 5.1, so the headline turnpike result does not follow from the transparent stabilizability/positivity assumptions alone. Section 6 attempts to discharge Assumption C under Assumption D, but the monotone-bounded-operator argument only yields strong convergence, not Hilbert–Schmidt norm convergence; the claimed existence of Λ∈L² and the L² limit are not rigorously established. This gap does not by itself constitute circularity, but it leaves the convergence as an assumption. The citation to [4] in Lemma 4.1 is not counted as circular because it concerns standard finite-dimensional Riccati estimates that are independent of the graphon structure; the self-citation is not the source of the circularity. Overall the paper is transparent about Assumption C, and the exponential-rate bootstrap is nontrivial, but a central part of the claimed contribution—the convergence of the Riccati system—is an input, meriting a score of 7.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No numerical free parameters are fitted; the coefficients are inputs. The main burden is carried by Assumptions C and D, which effectively import a large part of the asymptotic result into the hypotheses.

assumptions (5)
  • domain assumption Well-posedness of the graphon SDE (2) and measurability of u↦Law(...), taken from [20, Theorem 2.6]
    The paper builds its control formulation on this existence/uniqueness result.
  • domain assumption Assumption A: positivity of Q−S^T R^{-1}S, Q+Q̄−S^T R^{-1}S, Q̄, and positivity of kernel eQ G
    Used for coercivity of J_T and solvability of the finite-horizon Riccati equations.
  • domain assumption Assumption B: graphon-L² stabilizability of the mean ODE (18) and SDE (20)
    Yields exponential stability of the semigroup e^{tL} via Datko–Pazy; essential for ergodic solvability.
  • ad hoc to paper Assumption C: unique solvability of (24)-(25) and convergence Λ_T(t)→Λ for each fixed t as T→∞
    Contains the qualitative convergence that Prop 4.2 upgrades to exponential; central asymptotic content is assumed rather than derived.
  • ad hoc to paper Assumption D: operator matrix inequalities (48)-(49) involving O_i(t), O_i
    Given as sufficient condition for Assumption C; proof in Section 6 is sketched and only verified in special cases eC=0 or D=0.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Long-time behavior and turnpike properties of linear-quadratic graphon mean field control problems." pith.science (2026). https://pith.science/paper/USMMVRXD

@misc{pith2026260718000,
  author       = {Pith},
  title        = {Pith review of: Long-time behavior and turnpike properties of linear-quadratic graphon mean field control problems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/USMMVRXD}},
  note         = {Machine review of arXiv:2607.18000}
}
read the original abstract

We investigate the asymptotic behavior and turnpike properties of graphon mean field control (GMFC) problems in the linear-quadratic setting. We consider both a finite-horizon GMFC problem and its associated ergodic counterpart, in which the controlled dynamics are governed by a graphon mean field stochastic differential equation with heterogeneous interactions. The optimal controls and state trajectories for both problems are characterized by systems of Riccati equations together with systems of generalized differential and algebraic equations on suitable Hilbert spaces. Under a stabilizability condition and appropriate positivity assumptions on the graphon-induced operators, we establish the unique solvability of the ergodic control problem and derive exponential convergence estimates for the finite-horizon system to its stationary limit. As a consequence, we establish an exponential turnpike property for the optimal pair and prove the convergence of the time-averaged value function for the finite-horizon GMFC problem.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

44 extracted references · 4 linked inside Pith

  1. [1]

    Andersson and B

    D. Andersson and B. Djehiche. A maximum principle for SDEs of mean-field type.Applied Mathematics & Optimization, 63(3):341–356, 2011

    2011

  2. [2]

    Aurell, R

    A. Aurell, R. Carmona, and M. Lauri` ere. Stochastic graphon games: II. the linear-quadratic case. Applied Mathematics & Optimization, 85(3):39, 2022

    2022

  3. [3]

    Bayraktar, S

    E. Bayraktar, S. Chakraborty, and R. Wu. Graphon mean field systems.The Annals of Applied Probability, 33(5):3587–3619, 2023

    2023

  4. [4]

    Bayraktar and J

    E. Bayraktar and J. Jian. Ergodicity and turnpike properties of linear-quadratic mean field control problems.arXiv preprint arXiv:2502.08935, 2025

    arXiv 2025

  5. [5]

    Bayraktar and J

    E. Bayraktar and J. Jian. Convergence and turnpike properties of linear-quadratic mean field control problems with common noise.arXiv preprint arXiv:2601.07815, 2026

    arXiv 2026

  6. [6]

    Bayraktar, R

    E. Bayraktar, R. Wu, and X. Zhang. Propagation of chaos of forward–backward stochastic differential equations with graphon interactions.Applied Mathematics & Optimization, 88(1):25, 2023

    2023

  7. [7]

    Bayraktar and X

    E. Bayraktar and X. Zhang. Solvability of infinite horizon Mckean–Vlasov FBSDEs in mean field control problems and games.Applied Mathematics & Optimization, 87(1):13, 2023

    2023

  8. [8]

    Bensoussan, J

    A. Bensoussan, J. Frehse, P. Yam, et al.Mean field games and mean field type control theory, volume

Show all 44 references
  1. [9]

    Buckdahn, J

    R. Buckdahn, J. Li, and J. Ma. A stochastic maximum principle for general mean-field systems.Applied Mathematics & Optimization, 74(3):507–534, 2016

    2016

  2. [10]

    P. E. Caines, D. Ho, M. Huang, J. Jian, and Q. Song. On the graphon mean field game equations: Individual agent affine dynamics and mean field dependent performance functions.ESAIM: Control, Optimisation and Calculus of Variations, 28:24, 2022

    2022

  3. [11]

    P. E. Caines and M. Huang. Graphon mean field games and their equations.SIAM Journal on Control and Optimization, 59(6):4373–4399, 2021

    2021

  4. [12]

    Cao and M

    Z. Cao and M. Lauri` ere. Probabilistic analysis of graphon mean field control.arXiv preprint arXiv:2505.19664, 2025

    2025

  5. [13]

    Cardaliaguet, J.-M

    P. Cardaliaguet, J.-M. Lasry, P.-L. Lions, and A. Porretta. Long time average of mean field games with a nonlocal coupling.SIAM Journal on Control and Optimization, 51(5):3558–3591, 2013

    2013

  6. [14]

    Cardaliaguet and A

    P. Cardaliaguet and A. Porretta. Long time behavior of the master equation in mean field game theory. Analysis & PDE, 12(6):1397–1453, 2019

    2019

  7. [15]

    Carmona and F

    R. Carmona and F. Delarue.Probabilistic theory of mean field games with applications. I, volume 83 of Probability Theory and Stochastic Modelling. Springer, Cham, 2018. Mean field FBSDEs, control, and games

    2018

  8. [16]

    Cirant and A

    M. Cirant and A. Porretta. Long time behavior and turnpike solutions in mildly non-monotone mean field games.ESAIM: Control, Optimisation and Calculus of Variations, 27:86, 2021

    2021

  9. [17]

    Cohen and J

    A. Cohen and J. Jian. Turnpike properties in linear quadratic GaussianN-player differential games. ESAIM: Control, Optimisation and Calculus of Variations, 32:47, 2026

    2026

  10. [18]

    T. Damm, L. Gr¨ une, M. Stieler, and K. Worthmann. An exponential turnpike theorem for dissipative discrete time optimal control problems.SIAM Journal on Control and Optimization, 52(3):1935–1957, 2014

    1935

  11. [19]

    De Crescenzo, F

    A. De Crescenzo, F. de Feo, and H. Pham. Linear-quadratic optimal control for non-exchangeable mean-field SDEs and applications to systemic risk.ESAIM: Control, Optimisation and Calculus of Variations, 32:34, 2026

    2026

  12. [20]

    De Crescenzo, M

    A. De Crescenzo, M. Fuhrman, I. Kharroubi, and H. Pham. Mean-field control of non exchangeable systems.ESAIM: Control, Optimisation and Calculus of Variations, 32:3, 2026

    2026

  13. [21]

    Dunyak and P

    A. Dunyak and P. E. Caines. Quadratic optimal control of graphon Q-noise linear systems.Automatica, 190:113024, 2026

    2026

  14. [22]

    Engel and R

    K.-J. Engel and R. Nagel.One-parameter semigroups for linear evolution equations. Springer, 2000. LONG-TIME BEHA VIOR AND TURNPIKE PROPERTIES OF LQ GMFC PROBLEMS 43

    2000

  15. [23]

    Gao and P

    S. Gao and P. E. Caines. Optimal and approximate solutions to linear quadratic regulation of a class of graphon dynamical systems. In2019 IEEE 58th Conference on Decision and Control (CDC), pages 8359–8365. IEEE, 2019

    2019

  16. [24]

    Gugat, M

    M. Gugat, M. Herty, and C. Segala. The turnpike property for mean-field optimal control problems. European Journal of Applied Mathematics, 35(6):733–747, 2024

    2024

  17. [25]

    Huang, X

    J. Huang, X. Li, and J. Yong. A linear-quadratic optimal control problem for mean-field stochastic differential equations in infinite horizon.Mathematical Control and Related Fields, 5(1):97–139, 2015

    2015

  18. [26]

    J. Jian, S. Jin, Q. Song, and J. Yong. Long-time behaviors of stochastic linear-quadratic optimal control problems.Applied Mathematics & Optimization, 93(2):46, 2026

    2026

  19. [27]

    Kharroubi, S

    I. Kharroubi, S. Mekkaoui, and H. Pham. Stochastic maximum principle for optimal control problem of non exchangeable mean field systems.arXiv preprint arXiv:2506.05595, 2025

    2025 arXiv

  20. [28]

    M. Li, C. Mou, Z. Wu, and C. Zhou. Global well-posedness of Hamilton–Jacobi equations for linear- quadratic mean field control problems.Journal of Dynamics and Differential Equations, pages 1–36, 2025

    2025

  21. [29]

    X. Li, J. Sun, and J. Yong. Mean-field stochastic linear quadratic optimal control problems: closed-loop solvability.Probability, Uncertainty and Quantitative Risk, 1(1):2, 2016

    2016

  22. [30]

    Liang, B.-C

    Y. Liang, B.-C. Wang, and H. Zhang. Finite and infinite clusters mean field control problems via graphon theory. In2021 China Automation Congress (CAC), pages 5986–5990. IEEE, 2021

    2021

  23. [31]

    Lov´ asz.Large networks and graph limits, volume 60

    L. Lov´ asz.Large networks and graph limits, volume 60. American Mathematical Soc., 2012

    2012

  24. [32]

    H. Mei, R. Wang, and J. Yong. Turnpike property of stochastic linear-quadratic optimal control problems in large horizons with regime switching I: Homogeneous cases.arXiv preprint arXiv:2506.09337, 2025

    2025 arXiv

  25. [33]

    H. Mei, R. Wang, and J. Yong. Turnpike property of a linear-quadratic optimal control problem in large horizons with regime switching II: nonhomogeneous cases.ESAIM: Control, Optimisation and Calculus of Variations, 32:43, 2026

    2026

  26. [34]

    Nourian, P

    M. Nourian, P. E. Caines, R. P. Malham´ e, and M. Huang. Nash, social and centralized solutions to consensus problems via mean field control theory.IEEE Transactions on Automatic Control, 58(3):639– 653, 2012

    2012

  27. [35]

    Pham and X

    H. Pham and X. Wei. Dynamic programming for optimal control of stochastic Mckean–Vlasov dynamics. SIAM Journal on Control and Optimization, 55(2):1069–1101, 2017

    2017

  28. [36]

    Porretta and E

    A. Porretta and E. Zuazua. Long time versus steady state optimal control.SIAM Journal on Control and Optimization, 51(6):4242–4273, 2013

    2013

  29. [37]

    J. Sun. Mean-field stochastic linear quadratic optimal control problems: Open-loop solvabilities. ESAIM: Control, Optimisation and Calculus of Variations, 23(3):1099–1127, 2017

    2017

  30. [38]

    J. Sun, H. Wang, and J. Yong. Turnpike properties for stochastic linear-quadratic optimal control problems.Chinese Annals of Mathematics, Series B, 43(6):999–1022, 2022

    2022

  31. [39]

    Sun and J

    J. Sun and J. Yong. Turnpike properties for mean-field linear-quadratic optimal control problems.SIAM Journal on Control and Optimization, 62(1):752–775, 2024

    2024

  32. [40]

    Tr´ elat and E

    E. Tr´ elat and E. Zuazua. The turnpike property in finite-dimensional nonlinear optimal control.Journal of Differential Equations, 258(1):81–114, 2015

    2015

  33. [41]

    Tr´ elat and E

    E. Tr´ elat and E. Zuazua. Turnpike in optimal control and beyond: a survey.arXiv preprint arXiv:2503.20342, 2025

    2025 arXiv

  34. [42]

    Xiong and W

    J. Xiong and W. Xu. Mean-field stochastic linear quadratic control problem with random coefficients. SIAM Journal on Control and Optimization, 63(4):3042–3060, 2025

    2025

  35. [43]

    D. Xu, Z. Gou, and N. Huang. Social optima in linear quadratic graphon field control: analysis via infinite dimensional approach.arXiv preprint arXiv:2412.19082, 2024

    2024 arXiv

  36. [44]

    J. Yong. Linear-quadratic optimal control problems for mean-field stochastic differential equations. SIAM journal on Control and Optimization, 51(4):2809–2838, 2013. 44 E. BAYRAKTAR, Z. CAO, AND J. JIAN Department of Mathematics, University of Michigan, Ann Arbor, MI 48109, Un...

    2013

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.