Pith. sign in

REVIEW 2 major objections 4 minor 30 references

Zero-sum stochastic linear-quadratic games with Markov regime switching satisfy an exponential turnpike property.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

Finite-horizon optimal feedback gains in zero-sum stochastic linear-quadratic games with regime switching converge exponentially to infinite-horizon gains, yielding a turnpike theorem for the optimal triple.

T0 review reviewed 2026-08-04 challenge →

load-bearing objection Genuine extension with an exponential Riccati engine, but Theorem 5.2's proof has a real gap in the state-difference estimate; repairable, and worth refereeing. the 2 major comments →

arxiv 2509.09358 v1 pith:B6UJ2OJI submitted 2025-09-11 math.OC

Turnpike properties for zero-sum stochastic linear quadratic differential games of Markovian regime switching system

Xun Li , Fan Wu , Xin Zhang This is my paper
classification math.OC MSC 49N1091A2393E20
keywords zero-sum stochastic differential gameslinear-quadratic gamesturnpike propertycoupled differential Riccati equationscoupled algebraic Riccati equationsMarkov regime switchingopen-loop saddle strategyexponential convergence
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

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 proves an exponential turnpike property for zero-sum stochastic linear-quadratic (SLQ) differential games whose coefficients switch with a finite-state Markov chain. For a long finite horizon T, the unique open-loop saddle strategy of the game stays close to the stationary saddle strategy of the infinite-horizon problem, except near the two endpoints, and the distance is bounded explicitly by K times |x|^2 times (e^{-mu(T-t)} plus e^{-mu t}). The proof works by showing that the coupled differential Riccati equations of the finite-horizon problem converge exponentially, with the same rate, to the coupled algebraic Riccati equations of the infinite-horizon problem. If correct, the infinite-horizon stationary feedback can serve as a provably near-optimal strategy for long finite-horizon games, with an error that shrinks exponentially in the horizon length.

Core claim

Under assumptions (A1) and (A2), for every initial pair (x,i) there exist K, mu>0, independent of T, such that E[|X_T(t)-X_infty(t)|^2+|u_{1,T}(t)-u_{1,infty}(t)|^2+|u_{2,T}(t)-u_{2,infty}(t)|^2] <= K|x|^2(e^{-mu(T-t)}+e^{-mu t}), where the subscript T denotes the finite-horizon optimal triple and the subscript infinity the infinite-horizon optimal triple. The proof first shows that the unique strongly regular solution P_T of the coupled differential Riccati equations converges to the unique solution P_infty of the coupled algebraic Riccati equations at rate K e^{-mu(T-t)} (Theorem 5.1), and then transfers this convergence, via the closed-loop feedback representations, into the state and con

What carries the argument

Coupled differential Riccati equations (CDREs) for the finite-horizon game and their algebraic counterparts (CAREs) for the infinite-horizon game. The CDRE solution P_T(t,i) encodes the value function V_T(x,i)=<P_T(t,i)x,x> and yields the open-loop saddle strategy in closed-loop form, u_T(t)=Theta_T(t,alpha_t)X_T(t), with Theta_T(t,i)=-N(t;P_T,i)^{-1}L(t;P_T,i)^top. The core argument proves that P_T converges exponentially to the unique CARE solution P_infty and that the feedback gains converge at the same rate; the stable dynamics under the infinite-horizon feedback then convert this Riccati convergence into the turnpike estimate via a Lyapunov/Gronwall step. The uniform convexity-concavity

Load-bearing premise

The whole argument leans on (A1), which requires that, uniformly over every horizon and every Markov regime, Player 1's control cost is bounded below by delta times its squared norm and Player 2's payoff is bounded above by -delta times its squared norm; if this uniform convexity-concavity fails, the Riccati equations may have no strongly regular solution and the exponential turnpike proof collapses.

What would settle it

Take a one-dimensional, two-regime system satisfying (A2), set R11(1)=0 so that (A1) fails, and check whether a unique finite-horizon saddle still exists while the claimed exponential estimate fails. Under (A1)-(A2), a more direct falsifier is numerical: fix a two-regime example, solve the CDREs (20) backward, simulate the two closed-loop systems, and look for any t, T, x for which the bound (73) is exceeded; a single such counterexample would disprove Theorem 5.2.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • For any fixed kappa in (0,1/2), once T is large, the finite-horizon optimal state and controls on [kappa T,(1-kappa)T] are within O(e^{-mu kappa T}) of the infinite-horizon optimal triple, uniformly in the initial pair up to the factor |x|^2.
  • The infinite-horizon stationary feedback can be used as an approximate solution of the finite-horizon game, and inequality (4) gives an explicit, computable error estimate for that approximation.
  • The coupled algebraic Riccati equations (27) admit a unique solution satisfying the stabilizability condition (28), so the infinite-horizon problem is uniquely solvable and its value function is V_infty(x,i)=<P_infty(i)x,x>.
  • The value functions differ by an exponentially small horizon-truncation penalty: |V_T(x,i)-V_infty(x,i)| <= K|x|^2 e^{-mu T}, directly from the Riccati convergence rate at t=0.
  • The same exponential rate applies to the feedback gains, so the closed-loop approximation inherits the turnpike property.

Where Pith is reading between the lines

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

  • Editorial inference: the proof's structure suggests the same template of exponential Riccati convergence plus stability of the infinite-horizon closed loop would yield turnpike estimates for regime-switching games with periodic or ergodic coefficients, with the stationary CARE replaced by the corresponding periodic or ergodic Riccati solution; the paper does not claim this extension.
  • Because (A1) enters only through uniform invertibility of N, a testable extension is to let the convexity-concavity constant delta shrink with T and track how the turnpike rate depends on delta; the current theorem requires delta fixed.
  • The explicit constants K and mu in the turnpike bound depend on the Lyapunov exponent of the stable closed-loop system under Theta_infty; estimating them from data would let an engineer decide in advance whether a given horizon T is long enough for the infinite-horizon strategy to be used—an inference not drawn in the paper.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper studies the long-time behavior of zero-sum stochastic linear-quadratic differential games with Markov regime switching. Under a uniform convexity-concavity assumption (A1) and an L2-stability assumption (A2), it claims: (i) unique open-loop solvability of the finite- and infinite-horizon problems; (ii) convergence of the associated coupled differential Riccati equations (CDREs) to coupled algebraic Riccati equations (CAREs); (iii) exponential convergence of the Riccati solutions; and (iv) an exponential turnpike estimate for the optimal triple, i.e. the finite-horizon optimal state and controls are close to the infinite-horizon stationary feedback away from the temporal endpoints. The main tools are Hilbert-space operator representations of the cost, FBSDE stationarity conditions, Lyapunov/Gronwall arguments, and Riccati comparison.

Significance. If the main theorem is correct, this is a meaningful extension of the turnpike literature: it moves from deterministic or single-player stochastic LQ problems, and from zero-sum games without regime switching, to a Markovian regime-switching zero-sum SLQ setting, with explicit exponential rates. The paper also consolidates and extends the authors' previous solvability results for finite and infinite horizons. The operator-based uniqueness arguments and the Riccati convergence analysis are valuable in themselves. However, the proof of the central turnpike estimate (Theorem 5.2) contains a serious gap that must be repaired before the main claim can be regarded as established.

major comments (2)
  1. [§5, proof of Theorem 5.2 (Eq. (74))] The proof defines X_T = \bar X_T - \bar X_\infty, which satisfies the forced SDE (74) with X_T(0)=0. The text states that 'by (72) and Corollary 2.1, one has E|X_T(t)|^2 ≤ K|x|^2 e^{-\mu_3 t}'. Corollary 2.1 applies to a homogeneous L2-stable system with arbitrary initial condition, not to a forced process with zero initial condition. The claimed decay in t is essentially the turnpike estimate being proved, while the forcing term (\Theta_T-\Theta_\infty)\bar X_T is only known to decay in T-t. Without the e^{-\mu_3 t} bound, the line K|\Theta_\infty-\Theta_T| E[|\Sigma^{1/2}X_T||\bar X_T|] ≤ K|x|^2 e^{-2\mu_2(T-t)}e^{-\mu_3 t} is unsupported and the Gronwall step does not close. This is a load-bearing gap in the proof of (73).
  2. [§5, proof of Theorem 5.1 (Eqs. (66)-(69))] Equation (67) is stated as a linear bound |[\Theta-\Theta_T]^\top N_T[\Theta-\Theta_T]| ≤ K_1|\Sigma_T(t)|, but by (66) \Theta-\Theta_T is linear in \Sigma_T, so the left-hand side is quadratic in \Sigma_T. The subsequent inequality (69) uses |\Sigma_T(s)|^2, which is consistent only with a quadratic bound, not with the displayed (67). If (67) is taken literally, the comparison argument leading to h(t)≤\mu does not follow. Additionally, in Step 2 the replacement of |\Sigma_T(k)| by \rho is not explicitly justified; it can likely be obtained from Step 1 by choosing N large, but this needs to be written. Since (61) is used to derive (72) and then Theorem 5.2, these details matter.
minor comments (4)
  1. [§2, Proposition 2.1] Proposition 2.1 is stated under assumption (A1), but the estimate is a stability/regularity result for the linear SDE (9) and requires L2-stability of [A,C] (i.e., (A2)), not the uniform convexity-concavity condition. Corollary 2.1 inherits this mislabeling. Since (A2) is assumed in all main theorems, this is repairable, but the statement should be corrected.
  2. [§3, Theorem 3.3 proof] The system displayed after 'equation (43) is in turn equivalent to' repeats the first equation twice: the second line should read M^α_{21,∞}\bar u_{1,∞}+M^α_{22,∞}\bar u_{2,∞}+K^α_{2,∞}x=0.
  3. [§2, Eq. (39)] The formula for M^α_{ij,T} writes the adjoint term as (L^α_{j,T})^*S_i(α)^\top; for off-diagonal entries this should be (L^α_{i,T})^*S_j(α)^\top. As written, the off-diagonal blocks are not consistent with a self-adjoint M^α_T.
  4. [§4, proof of Theorem 4.1] The line '||\hat u_T(·)||≤KE|Y_∞(T)|^2' mixes the Hilbert-space norm with the expectation. It should read E||\hat u_T(·)||^2 ≤ K E|Y_∞(T)|^2.

Circularity Check

0 steps flagged

No definitional or self-citation circularity; the central turnpike proof is independent of the claimed conclusion, though it contains a non-circular proof gap in Theorem 5.2.

full rationale

Walking the derivation: (A1)-(A2) are assumptions, not the turnpike inequality (73). Theorems 3.1/3.2, Corollary 2.1, and Proposition 2.1 are imported from the authors' prior works [27,28,29], but those works establish solvability, FBSDE characterizations and stability estimates, not the exponential turnpike bound. Theorem 5.1 derives Riccati convergence from the CDREs by a Gronwall argument; Theorem 5.2 then uses (72) and a Lyapunov differential inequality for E<Sigma(alpha)X_T,X_T>. The target (73) does not appear among the inputs of any lemma; the proof is therefore not circular in the self-definitional or fitted-parameter sense. The one serious issue is the sentence in the proof of Theorem 5.2: 'by (72) and Corollary 2.1, one has ... E|X_T(t)|^2 <= K|x|^2 e^{-mu3 t}'. Corollary 2.1 applies to the homogeneous system [A,C] with arbitrary initial condition, whereas X_T is defined by (74) with X_T(0)=0 and nontrivial forcing terms. So the exponential-in-t bound on X_T is not supplied by the cited result; it is a proof gap. This is a correctness concern, not a circular reduction: the asserted bound is not the theorem's conclusion, is not a hidden restatement of an assumption, and is not shown to be equivalent to the quantity being estimated. Accordingly, the circularity score is 0.

Axiom & Free-Parameter Ledger

0 free parameters · 4 axioms · 0 invented entities

No free parameters are fitted. The central claim rests on structural assumptions A1 and A2 plus stability and representation results imported from the authors' prior papers [27,28,29]. No new entities such as particles, forces, or dimensions are postulated.

axioms (4)
  • domain assumption A1: uniform convexity-concavity (10) for all T in (0,infinity] and all i in S
    Ensures N11 and -N22 blocks are uniformly sign-definite, M operators are invertible with bounded inverses, and CDRE solutions are strongly regular. Used in Theorems 3.1, 3.3, 4.1, 5.1 and 5.2.
  • domain assumption A2: L2-stability of the uncontrolled system [A,C]_alpha
    Makes the infinite-horizon performance functional well-defined on all of U1 x U2 and gives exponential stability estimates. Used in Theorem 3.2 and throughout Sections 4 and 5.
  • standard math Finite-state irreducible Markov chain alpha with generator Pi, independent of a one-dimensional Brownian motion
    Background probabilistic setup from Section 1; needed for martingale representation of jump terms in the FBSDEs and for the regime-switching dynamics.
  • domain assumption Stability and Lyapunov estimates from Wu-Li-Zhang [29, Prop. 2.2 and 2.5]
    Imported to bound solutions of affine SDEs and to obtain the Lyapunov inequality used in Theorem 5.2. This is a cited result from the authors' own prior work.

reviewed 2026-08-04 · how reviews work

0 comments
Cite this review

Pith. "Pith review of Turnpike properties for zero-sum stochastic linear quadratic differential games of Markovian regime switching system." pith.science (2026). https://pith.science/paper/B6UJ2OJI

@misc{pith2026250909358,
  author       = {Pith},
  title        = {Pith review of: Turnpike properties for zero-sum stochastic linear quadratic differential games of Markovian regime switching system},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/B6UJ2OJI}},
  note         = {Machine review of arXiv:2509.09358}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

This paper investigates the long-time behavior of zero-sum stochastic linear-quadratic (SLQ) differential games within Markov regime-switching diffusion systems and establishes the turnpike property of the optimal triple. By verifying the convergence of the associated coupled differential Riccati equations (CDREs) along with their convergence rate, we show that, for a sufficiently large time horizon, the equilibrium strategy in the finite-horizon problem can be closely approximated by that of the infinite-horizon problem. Furthermore, this study enhances and extends existing results concerning zero-sum SLQ differential games over both finite and infinite horizons.

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

30 extracted references · 4 linked inside Pith

  1. [1]

    A weak dynamic programming principle for zero-sum stochastic differential games with unbounded controls.SIAM Journal on Control and Optimization, 51(3): 2036–2080, 2013

    Erhan Bayraktar and Song Yao. A weak dynamic programming principle for zero-sum stochastic differential games with unbounded controls.SIAM Journal on Control and Optimization, 51(3): 2036–2080, 2013

    2036

  2. [2]

    On the turnpike property and the receding-horizon method for linear-quadratic optimal control problems.SIAM Journal on Control and Optimization, 58(2): 1077–1102, 2020

    Tobias Breiten and Laurent Pfeiffer. On the turnpike property and the receding-horizon method for linear-quadratic optimal control problems.SIAM Journal on Control and Optimization, 58(2): 1077–1102, 2020

    2020

  3. [3]

    Stochastic differential games and viscosity solutions of Hamilton- Jacobi-Bellman-Isaacs equations.SIAM Journal on Control and Optimization, 47(1):444–475, 2008

    Rainer Buckdahn and Juan Li. Stochastic differential games and viscosity solutions of Hamilton- Jacobi-Bellman-Isaacs equations.SIAM Journal on Control and Optimization, 47(1):444–475, 2008

    2008

  4. [4]

    An exponential turnpike theorem for dissipative discrete time optimal control problems.SIAM Journal on Control and Optimization, 52(3):1935–1957, 2014

    Tobias Damm, Lars Grune, Marleen Stieler, and Karl Worthmann. An exponential turnpike theorem for dissipative discrete time optimal control problems.SIAM Journal on Control and Optimization, 52(3):1935–1957, 2014

    1935

  5. [5]

    McGraw-Hill, New York, 1958

    Robert Dorfman, Paul Anthony Samuelson, and Robert M Solow.Linear Programming and Eco- nomic Analysis. McGraw-Hill, New York, 1958

    1958

  6. [6]

    On the existence of value functions of two-player, zero-sum stochastic differential games.Indiana University Mathematics Journal, 38(2):293–314, 1989

    Wendell H Fleming and Panagiotis E Souganidis. On the existence of value functions of two-player, zero-sum stochastic differential games.Indiana University Mathematics Journal, 38(2):293–314, 1989

    1989

  7. [7]

    Turnpike properties and strict dissipativity for discrete time linear quadratic optimal control problems.SIAM Journal on Control and Optimization, 56(2):1282– 1302, 2018

    Lars Gr¨ une and Roberto Guglielmi. Turnpike properties and strict dissipativity for discrete time linear quadratic optimal control problems.SIAM Journal on Control and Optimization, 56(2):1282– 1302, 2018

    2018

  8. [8]

    Zero-sum stochastic differential games and backward equations.Systems & Control Letters, 24(4):259–263, 1995

    Saıd Hamad´ ene and Jean-Pierre Lepeltier. Zero-sum stochastic differential games and backward equations.Systems & Control Letters, 24(4):259–263, 1995

    1995

  9. [9]

    Turnpike properties of optimal relaxed control problems.ESAIM: Control, Optimisation and Calculus of Variations, 25:74, 2019

    Hongwei Lou and Weihan Wang. Turnpike properties of optimal relaxed control problems.ESAIM: Control, Optimisation and Calculus of Variations, 25:74, 2019

    2019

  10. [10]

    Two-player zero-sum stochastic differential games with regime switching.Automatica, 114: 108819, 2020

    Siyu Lv. Two-player zero-sum stochastic differential games with regime switching.Automatica, 114: 108819, 2020. 22

    2020

  11. [11]

    Turnpike property of stochastic linear-quadratic optimal control problems in large horizons with regime switching i: Homogeneous cases.arXiv preprint arXiv:2506.09337, 2025

    Hongwei Mei, Rui Wang, and Jiongmin 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

    Pith/arXiv arXiv 2025

  12. [12]

    Two-person zero-sum linear quadratic stochastic differential games by a Hilbert space method.Journal of Industrial and Management Optimization, 2(1):93–115, 2006

    Libin Mou, Jiongmin Yong, et al. Two-person zero-sum linear quadratic stochastic differential games by a Hilbert space method.Journal of Industrial and Management Optimization, 2(1):93–115, 2006

    2006

  13. [13]

    A model of general economic equilibrium.The Review of Economic Studies, 13(1): 1–9, 1945

    J v Neumann. A model of general economic equilibrium.The Review of Economic Studies, 13(1): 1–9, 1945

    1945

  14. [14]

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

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

    2013

  15. [15]

    A mathematical theory of saving.The economic journal, 38(152):543–559, 1928

    Frank Plumpton Ramsey. A mathematical theory of saving.The economic journal, 38(152):543–559, 1928

    1928

  16. [16]

    The turnpike property in nonlinear optimal control—a geometric approach.Automatica, 134:109939, 2021

    Noboru Sakamoto and Enrique Zuazua. The turnpike property in nonlinear optimal control—a geometric approach.Automatica, 134:109939, 2021

    2021

  17. [17]

    Turnpike and dissipativity in generalized discrete-time stochastic linear-quadratic optimal control.SIAM Journal on Control and Optimization, 63(2):1432–1457, 2025

    Jonas Schießl, Ruchuan Ou, Timm Faulwasser, Michael H Baumann, and Lars Gr¨ une. Turnpike and dissipativity in generalized discrete-time stochastic linear-quadratic optimal control.SIAM Journal on Control and Optimization, 63(2):1432–1457, 2025

    2025

  18. [18]

    Two-person zero-sum stochastic linear-quadratic differential games.SIAM Journal on Control and Optimization, 59(3):1804–1829, 2021

    Jingrui Sun. Two-person zero-sum stochastic linear-quadratic differential games.SIAM Journal on Control and Optimization, 59(3):1804–1829, 2021

    2021

  19. [19]

    Linear quadratic stochastic differential games: Open-loop and closed-loop saddle points.SIAM Journal on Control and Optimization, 52(6):4082–4121, 2014

    Jingrui Sun and Jiongmin Yong. Linear quadratic stochastic differential games: Open-loop and closed-loop saddle points.SIAM Journal on Control and Optimization, 52(6):4082–4121, 2014

    2014

  20. [20]

    Long-time behavior of zero-sum linear-quadratic stochastic differ- ential games.arXiv preprint arXiv:2406.02089, 2024

    Jingrui Sun and Jiongmin Yong. Long-time behavior of zero-sum linear-quadratic stochastic differ- ential games.arXiv preprint arXiv:2406.02089, 2024

    Pith/arXiv arXiv 2024

  21. [21]

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

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

    2024

  22. [22]

    Turnpike properties for stochastic linear-quadratic optimal control problems with periodic coefficients.Journal of Differential Equations, 400:189–229, 2024

    Jingrui Sun and Jiongmin Yong. Turnpike properties for stochastic linear-quadratic optimal control problems with periodic coefficients.Journal of Differential Equations, 400:189–229, 2024

    2024

  23. [23]

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

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

    2022

  24. [24]

    The exponential turnpike property for periodic linear quadratic optimal control problems in infinite dimension.SIAM Journal on Control and Optimization, 63(4):2524–2546, 2025

    Emmanuel Tr´ elat, Xingwu Zeng, and Can Zhang. The exponential turnpike property for periodic linear quadratic optimal control problems in infinite dimension.SIAM Journal on Control and Optimization, 63(4):2524–2546, 2025

    2025

  25. [25]

    A pontryagin’s maximum principle for non-zero sum differential games of BSDEs with applications.IEEE Transactions on Automatic control, 55(7):1742–1747, 2010

    Guangchen Wang and Zhiyong Yu. A pontryagin’s maximum principle for non-zero sum differential games of BSDEs with applications.IEEE Transactions on Automatic control, 55(7):1742–1747, 2010

    2010

  26. [26]

    A partial information non-zero sum differential game of backward stochastic differential equations with applications.Automatica, 48(2):342–352, 2012

    Guangchen Wang and Zhiyong Yu. A partial information non-zero sum differential game of backward stochastic differential equations with applications.Automatica, 48(2):342–352, 2012

    2012

  27. [27]

    Stochastic linear-quadratic differential game with Markovian jumps in an infinite horizon.arXiv preprint arXiv:2408.12818, 2024

    Fan Wu, Xun Li, Jie Xiong, and Xin Zhang. Stochastic linear-quadratic differential game with Markovian jumps in an infinite horizon.arXiv preprint arXiv:2408.12818, 2024. 23

    Pith/arXiv arXiv 2024

  28. [28]

    Open-loop and closed-loop solvabilities for zero-sum stochas- tic linear quadratic differential games of Markovian regime switching system.arXiv preprint arXiv:2409.01973, 2024

    Fan Wu, Xun Li, and Xin Zhang. Open-loop and closed-loop solvabilities for zero-sum stochas- tic linear quadratic differential games of Markovian regime switching system.arXiv preprint arXiv:2409.01973, 2024

    Pith/arXiv arXiv 2024

  29. [29]

    Stochastic linear quadratic optimal control problems with regime- switching jumps in infinite horizon.SIAM Journal on Control and Optimization, 63(2):852–891, 2025

    Fan Wu, Xun Li, and Xin Zhang. Stochastic linear quadratic optimal control problems with regime- switching jumps in infinite horizon.SIAM Journal on Control and Optimization, 63(2):852–891, 2025

    2025

  30. [30]

    Zhiyong Yu. An optimal feedback control-strategy pair for zero-sum linear-quadratic stochastic differential game: The Riccati equation approach.SIAM Journal on Control and Optimization, 53 (4):2141–2167, 2015. 24

    2015

This paper was first reviewed by deepseek-v4-flash on August 4, 2026.