Pith. sign in

REVIEW 5 major objections 5 minor 46 references

Mean-Field-Type Game Theory with Rosenblatt Noise

T0 review · 5 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Rosenblatt noise admits closed-form optimal controls and game equilibria.

desk verdict The variance-aware control and renewable-energy game results are genuinely new, but Theorem 2's prediction claim is wrong and the MFTG existence theorems are conditional on unproven ODE solvability. read the letter →

arxiv 2506.08025 v1 pith:Q7BHVHAB submitted 2025-05-29 math.OC cs.GT

Hamidou Tembine , Tyrone E. Duncan , Bozenna Pasik-Duncan This is my paper
classification math.OCcs.GT MSC 49N1060G2260H0591A1591A1693E20
keywords Rosenblattprocessmean-field-typegamesnon-Gaussiannoiselong-rangedependencelinear-quadraticcontrolNashequilibriumsaddle-pointstochasticcalculus
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 builds a control and game-theoretic framework in which the randomness is a Rosenblatt process, a non-Gaussian, self-similar, long-range-dependent noise that is not a semimartingale. It uses a stochastic calculus for this noise to solve a linear-quadratic control problem in closed form, giving the optimal feedback gain explicitly in terms of the Hurst parameter $H$. The same machinery yields saddle-point strategies for zero-sum games and state-feedback Nash equilibria for non-zero-sum mean-field-type games, provided certain Riccati differential equations admit positive non-exploding solutions. The paper's central warning is that approximating Rosenblatt noise by Brownian motion, fractional Brownian motion, or other standard noises makes controls strictly suboptimal and turns equilibrium strategies into non-equilibria. This matters because the motivating applications, such as power grids, e-commerce, agriculture, and underwater channels, exhibit the skew, heavy tails, and long memory that Gaussian models miss.

What carries the argument

The central object is the Rosenblatt process $R_H(t)$, defined by a double Wiener-Itô integral and normalized so that $\mathbb{E}[(R_H(t))^2]=1$; it is self-similar with Hurst parameter $H\in(\frac12,1)$, non-Gaussian, long-range dependent, and not a semimartingale. The load-bearing tool is the stochastic calculus formula from the paper's reference [27], which expresses $f(t,x(t))$ for Rosenblatt-driven $x$ as a sum of a Lebesgue integral, a fractional Brownian integral, and a Rosenblatt integral, with a third derivative of $f$ appearing in the drift. That formula lets the authors compute $\mathbb{E}[X_T^2]$ explicitly via a fractional derivative $\nabla^{H/2,H/2}X_t(t,t)$ and Gamma-function asymptotics, reducing the control problem to a quadratic equation in the gain $K$. In the games, an orthogonal decomposition $x=\tilde{x}+\bar{x}$ separates the Rosenblatt-driven fluctuations from the deterministic mean-field dynamics, which makes the mean-field-type game semi-explicitly solvable by a direct method rather than by dynamic programming.

What would settle it

Choose concrete coefficients for a two-player non-zero-sum game and solve the coupled Riccati system numerically over the horizon $[0,T]$: if the solution blows up before $T$, the claimed Nash equilibrium does not exist. For the linear-quadratic control claim, simulate the Rosenblatt-driven state under the optimal gain $\hat{K}^*$ and under the gain obtained by first approximating the noise as Brownian motion; if the approximate controller does not have strictly larger long-run average cost, the suboptimality claim is falsified.

Watch

Extended reading notes

Core claim

For the infinite-horizon linear-quadratic control problem with state $x(t)=x_0+\int_0^t (b_1 x(s)+b_2 u(s))\,ds+R_H(t)$, the paper claims the optimal linear state-feedback gain is $\hat{K}^* = -\frac{b_1 + \sqrt{b_1^2 + 4H(1-H)b_2^2 q/r}}{2 b_2 (1-H)}$, with long-run average cost $L_\infty(\hat{K}^*) = \frac{\Gamma(2H+1)(q+r(\hat{K}^*)^2)}{2[-(b_1+b_2\hat{K}^*)]^{2H}}$. The paper further claims that replacing the Rosenblatt noise by Brownian motion, fractional Brownian motion, Gauss-Volterra noise, or a compensated Poisson jump process and then optimizing leads to suboptimal control. In zero-sum games it claims a saddle-point value exists with explicit gains, and in non-zero-sum and mean-field-type games it gives closed-form stationary linear state-feedback Nash equilibria whenever a coupled system of Riccati differential equations has a positive solution that does not blow up on the horizon. For a renewable-energy producer game, the equilibrium separates into a Rosenblatt-driven fluctuating component and a deterministic mean-field Cournot component, with the fluctuating gain depending explicitly on $H$.

Load-bearing premise

The game-theoretic results are conditional on a coupled system of Riccati differential equations having a positive solution that exists over the whole planning horizon without blowing up, and the paper does not prove such a solution always exists.

Editorial extensions

If this is right

  • For the linear-quadratic control problem, the optimal gain is $\hat{K}^* = -\frac{b_1 + \sqrt{b_1^2 + 4H(1-H)b_2^2 q/r}}{2 b_2 (1-H)}$, and it reduces to the classical LQ gain only in the limit where the Hurst parameter degenerates.
  • Any approximation of Rosenblatt noise by Brownian motion, fractional Brownian motion, Gauss-Volterra noise, or compensated Poisson jumps yields a strictly suboptimal controller, and the suboptimality persists for variance-aware costs.
  • Zero-sum games driven by Rosenblatt noise have a value, and the saddle-point gains solve an explicit quadratic equation that depends on $H$.
  • Non-zero-sum games and mean-field-type games have stationary linear state-feedback Nash equilibria in closed form whenever the coupled Riccati differential equations admit a positive solution that does not blow up.
  • In the renewable-energy producer example, the equilibrium production strategy splits into a Rosenblatt-driven term with gain depending on $H$ and a deterministic mean-field Cournot term, showing that variance-awareness changes the equilibrium itself.

Reading between the lines

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

  • An implication the paper leaves implicit is that the Rosenblatt suboptimality results give modelers a practical rule: estimate the Hurst parameter of the noise before choosing a controller, since replacing long-memory noise by Gaussian noise is not merely a small perturbation but changes the optimal policy.
  • The paper sketches Rosenblatt-driven diffusion models but does not report generative experiments; a testable extension would be to train a diffusion transformer on data with known long-range dependence and compare samples and likelihoods against the Brownian version.
  • If the Riccati existence condition fails for a concrete game instance, the formulas in the theorems define formal candidates rather than actual equilibria; numerically mapping where blow-up occurs would delimit the theory's scope.
  • The same Rosenblatt stochastic calculus could likely be applied to filtering and estimation problems, not just prediction and control, although the paper does not develop filters.
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

5 major / 5 minor

Summary. The manuscript develops a framework for stochastic control and mean-field-type games driven by the Rosenblatt process, a non-Gaussian, self-similar, long-range-dependent process. It presents a stochastic calculus rule (Theorem 1, taken from [27]), prediction formulas (Theorems 2–3), an infinite-horizon linear-quadratic optimal control result with explicit feedback gain (Theorem 4), zero-sum game saddle-point results (Theorems 6–7), non-zero-sum game best-response and equilibrium conditions (Theorems 8–9), and mean-field-type game Nash equilibria and cooperative solutions stated as conditional on the solvability of coupled Riccati systems (Theorems 10–11). It then applies the framework to a Cournot mean-field-type game (Theorems 12–18) and compares mean-field-type games with multi-population mean-field games. The paper also contains extensive empirical motivation and a section on Rosenblatt-based diffusion transformers.

Significance. If the explicit formulas and equilibrium characterizations are correct, they would provide tractable benchmarks for a class of genuinely non-Gaussian, long-memory stochastic control and game problems, for which standard Brownian or fractional-Brownian approximations are known to be inadequate. The paper's main value, however, rests on previously established results by the same group: Theorem 1 is taken from [27], Theorem 4 matches [43], and Theorems 6–7 match [44]. The genuinely new game-theoretic claims (Theorems 10–11) are conditional on unproven solvability of coupled ODE systems. The paper does not provide machine-checked proofs, reproducible code, or quantitative validation of its central existence claims; its empirical section makes strong quantitative claims (e.g., 60% tail-risk underestimation, Wasserstein error 10^-6) without data or statistical methodology. The framework is potentially useful for benchmarking and for motivating further work, but the central novelty is partially a repackaging of prior contributions.

major comments (5)
  1. [Section V, Theorem 2] Equation (4) asserts that E[x(t) | x(t''), 0<t''<t'] = x(t') for the solution of dX(t)=b1 X(t)dt+dRH(t). This is the martingale property. It is false for b1 ≠ 0 and for any H∈(1/2,1): the drift makes the conditional expectation of X(t) given the past a nontrivial forecast, not the current value, and the Rosenblatt process itself is not a martingale. Even for b1=0, the stochastic integral over (t',t] is generally not independent of the past because Rosenblatt increments are not independent. The claimed 'optimal mean square error prediction' is therefore incorrect as stated, and this is a load-bearing error in the prediction section.
  2. [Section XI, Theorems 10 and 11] The Nash equilibrium and cooperative optimality are stated as 'whenever' the displayed system of differential equations admits a positive solution that does not blow up on [0,T]. No existence or uniqueness conditions are proved, and no numerical demonstration is given. In particular, the coupled equations for \bar λ_i involve terms (b2j+\bar b2j)^{1/(2\bar k_j-1)} whose signs and growth properties depend on the parameters; for \bar k_j>1 and mixed signs, finite-time blow-up is not ruled out. Since the paper's central claim is the identification of state-feedback Nash equilibria, the absence of any solvability result leaves the existence claim unestablished. The conditional phrasing is honest, but the abstract's statement that the paper 'establishes conditions' and 'identifies' equilibria overstates what is shown.
  3. [Sections IX and X, Theorems 6–9] The zero-sum value and saddle-point gains (Theorems 6–7) and the non-zero-sum best-response/equilibrium results (Theorems 8–9) are stated without proof. Theorem 6 is a bare assertion; Theorem 7 is credited only implicitly to [44]; Theorem 8 is asserted with no derivation; and Theorem 9 is again conditional on the existence of a solution to a coupled algebraic system, with no fixed-point or monotonicity argument. For a self-contained journal submission, these central game-theoretic claims need either full proofs or precise, quoted statements from the cited references, and the existence conditions need to be spelled out and verified.
  4. [Section VII] The section titled 'Adaptive Control Driven by Rosenblatt Noise' is empty: the heading is followed immediately by Section VIII with no content, no equations, and no text. This is not a formatting nit; it indicates an incomplete manuscript and is a serious presentation defect for a journal submission.
  5. [Sections VI.A, IX.A, X.A, VIII.A] The paper repeatedly claims that replacing Rosenblatt noise by Brownian motion, fractional Brownian motion, Gauss–Volterra noise, or compensated Poisson noise leads to suboptimal control, non-equilibrium, or loss of the saddle point, and states '∃ϵ>0: |C_surrogate − C_true| ≥ ϵ' with no proof or quantification. These are central motivations for the paper, yet no theorem or numerical experiment substantiates them. If these claims are meant to be theorems, they need precise statements and proofs; otherwise they should be presented as conjectures or illustrated only informally.
minor comments (5)
  1. [Section VI, proof of Theorem 4] The displayed equation '0 = 2(b1+b2K)(q+rK^2)L∞(K) + Γ(2H+1)/[-(b1+b2K)]^{2H-1}' is dimensionally inconsistent with the preceding derivation and with the final formula for L∞(K). The first term should be 2(b1+b2K)L∞(K)/(q+rK^2), or an equivalent correction is needed.
  2. [Section IV and Theorem 10] Several constants are undefined or inconsistently used: c3 appears in the definition of o(t) without introduction; the relation between c, \tilde c, and \tilde C_H is not fully specified; and the notation limits 'lim infu,T→∞' in Section IX is ambiguous.
  3. [Section V] There is a typo 'Leet h be a positive continuous function' that should read 'Let h...'. The paragraph following Theorem 3 also contains undefined notation such as \hat g and \hat g^{exp} in the displayed formulas.
  4. [Section II] The empirical claims include specific quantitative assertions (e.g., 'Gaussian models underestimate tail risk by 60%', 'Wasserstein error on the order of 10^{-6}', 'mispredicted grid stress events by 40%') without data sources, sample sizes, or statistical methodology. If these are meant to support the motivation, references or a description of the data and fitting procedure are needed; otherwise they should be marked as illustrative.
  5. [Table IV and Section XIII] The statement that multi-population MFG is 'irrelevant, inefficient and suboptimal' for the variance-awareness problem is asserted with an 'unbounded price of simplicity' argument, but the dynamic extension with Rosenblatt noise is not formally proved; this should be stated more cautiously or given a proof.

Circularity Check

2 steps flagged · score 4.0 of 10

The Rosenblatt calculus and the zero-sum game value are imported from the authors' own prior papers, but Theorem 4 is a genuine optimization derivation; no fitted input is renamed as a prediction.

  1. self citation load bearing [Section IV, Theorem 1 and equation (3)]
    "This is a new Stochastic Calculus established in [27]."

    All later control and game derivations rely on this Itô formula. The proof of Theorem 4 explicitly says 'we use Theorem 1 and Equation (1)' to compute the evolution of the squared state. The theorem is not proved in the present paper; it is imported from [27], whose authors include two of the present authors. The abstract's contribution claim that 'We develop novel stochastic calculus formulas for a range of Rosenblatt processes' is therefore carried by a self-citation, and every subsequent formula inherits that dependence. Since the calculus is the foundation for the control and game results, this is load-bearing self-citation rather than a mere bibliographic mention, although [27] is a published theorem.

  2. self citation load bearing [Section IX, Theorems 6-7]
    "subject to the state dynamics driven by Rosenblatt noise [44]. ... Theorem 6. MiniMax=Maximin: The zero-sum game under Rosenblatt has a value and each decision-maker has an optimal strategy."

    The paper's zero-sum-game contribution, namely existence of a value and explicit saddle-point gains, is stated without proof and the problem is attributed to [44], a prior paper by two of the present authors. Unlike Theorem 4, no derivation is supplied in this text: Theorem 6 and Theorem 7 are simply asserted after the citation. Thus the central adversarial-game claim reduces to a self-citation, with the formulas in Theorem 7 presented as the outcome of that cited prior work rather than as a result derived within this paper.

full rationale

The paper does not exhibit the strongest forms of circularity: there are no fitted parameters renamed as predictions, no target quantities used in their own definitions, and no ansatz smuggled in by citation: the restriction to linear, mean-field-type feedback is explicitly stated as a restriction, with the extension to nonlinear strategies flagged as open. Theorem 4 is a genuine minimization over the one-dimensional gain K, obtained from a variation-of-constants representation together with the cited Itô formula; the optimal-cost identity follows from the quadratic equation for the gain rather than being assumed. Theorems 8, 12-18 solve explicit fixed-point or best-response conditions. The conditional Riccati solvability in Theorems 9-11 is stated honestly with 'whenever', but no existence proof is given; that is a mathematical-completeness gap, not a circular derivation. The main circularity concern is the load-bearing self-citation chain: the Rosenblatt stochastic calculus (Theorem 1) is taken from [27], and the zero-sum game value (Theorems 6-7) is taken from [44], both by the present authors, while the abstract claims these as new developments. That warrants a moderate score of 4; the independent optimization content in the control and nonzero-sum derivations prevents a higher score.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central math relies on the stochastic calculus of [27] as a background theorem. The empirical sections fit Rosenblatt shapes to data without specifying parameters or error bars. No new physical entities are introduced.

free parameters (2)
  • Hurst exponent H (underwater channel) = ≈0.8
    Reported as estimated from field data in Section II.H, but no dataset or estimation procedure is provided.
  • Rosenblatt shape parameters for power grid and mango data = not specified
    The paper claims data follow two-mode or three-mode Rosenblatt distributions but gives no parameter values or fitting methodology.
assumptions (4)
  • standard math Rosenblatt stochastic calculus (Theorem 1 of [27])
    Invoked as a black box in the proofs of Theorems 4, 5, 10, and subsequent game results.
  • standard math Stochastic integrals with respect to Rosenblatt process have zero expectation
    Used in the proof of Theorem 4 to drop integrated noise terms when taking expectations.
  • ad hoc to paper Existence of positive non-blow-up solution to the coupled Riccati system in Theorems 9-11
    The equilibrium results are conditional on this unproven existence.
  • domain assumption Orthogonal decomposition of state into deviation and mean-field components
    Used in Section XI.A to split the best-response problem into two independent subproblems.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mean-Field-Type Game Theory with Rosenblatt Noise." pith.science (2026). https://pith.science/paper/Q7BHVHAB

@misc{pith2026250608025,
  author       = {Pith},
  title        = {Pith review of: Mean-Field-Type Game Theory with Rosenblatt Noise},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q7BHVHAB}},
  note         = {Machine review of arXiv:2506.08025}
}
read the original abstract

We study the integration of Rosenblatt noise into stochastic systems, control theory, and mean-field-type game theory, addressing the limitations of traditional Gaussian and Markovian models. Empirical evidence from various domains, including water demand, e-commerce, power grid operations, wireless channels, and agricultural supply chains, demonstrates the prevalence of non-Gaussian characteristics such as skews, heavy tails and strong long-range dependencies. The Rosenblatt process, a non-Gaussian non-Markovian, self-similar process, offers a baseline framework for capturing some the behaviors observed in real data. We develop novel stochastic calculus formulas for Rosenblatt processes, apply these to dynamical systems, and analyze optimal control problems, revealing the suboptimality of traditional noise approximation methods. We extend game-theoretic analysis to environments driven by Rosenblatt noise, establishing conditions for saddle-point equilibria in zero-sum games and identifying state-feedback Nash equilibria in non-zero-sum games. Our findings underscore the importance of incorporating non-Gaussian noise into predictive analytics and control strategies, enhancing the accuracy and robustness of models in real-world applications. These findings represent a significant advancement in mean-field-type game theory with variance-awareness, offering new insights and tools for managing interactive systems influenced by Rosenblatt noise.

Figures

Figures reproduced from arXiv: 2506.08025 by the authors.

Figure 1
Figure 1. Top 5% ”whales” dominate revenue. Gaussian models underestimate tail risk by 60%. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. The influx of low-cost purchases accentuates the structure of the spending distribution making it more skewed. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Bimodal peaks (solar surplus vs evening demand) make Gaussian load forecasting error-prone. Net load= grid demand-local [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (14 more)
Figure 4
Figure 4. Figure 4: Net load: change in the total demand due to renewable energy sources during the sunny days and the windy days modelled using [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Daily net load during the summer on pumping to prevent seedling failure, creating a second, even larger peak. After the monsoon season (October-November), wheat planting begins, but depleted reservoirs force reliance on groundwater, driving the third peak. These three …
Figure 6
Figure 6. Figure 6: Trimodal Water Demand in Punjab (averaged Annual Total: 30 BCM). [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Bimodal Mango Supply Dual harvest seasons create 2-month scarcity gap - invisible to single-peak models. Over an average of 75 [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Technology: CPU/GPU/TPU Adoption. Demand spikes driven by crypto (2017, 2021, 2024) and LLMs (2023) violate Gaussian [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: Price index over the years. Crypto peaks (2017,2021, 2024) and LLM peak (2023) create non-Gaussian price shocks. [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]
Figure 10
Figure 10. Figure 10: Machine Intelligence Funding: Several peaks/waves/cycles: Machine Intelligence winters (1980s) vs GPT-era hypergrowth defy [PITH_FULL_IMAGE:figures/full_fig_p015_10.png]
Figure 11
Figure 11. Figure 11: The Gaussian fit does not capture the skewed and heavy-tailed behavior of the actual data. [PITH_FULL_IMAGE:figures/full_fig_p016_11.png]
Figure 12
Figure 12. Figure 12: Non-Gaussian fluctuations in shallot and onion prices over time. [PITH_FULL_IMAGE:figures/full_fig_p016_12.png]
Figure 13
Figure 13. Figure 13: Symmetry does not necessarily imply Gaussianity [PITH_FULL_IMAGE:figures/full_fig_p017_13.png]
Figure 14
Figure 14. Figure 14: Measured channel impulse response recorded under deep-water conditions at a transmitter depth of 20m. [PITH_FULL_IMAGE:figures/full_fig_p018_14.png]
Figure 15
Figure 15. Figure 15: Probability density function of SoH for EV batteries in 5+ year-old electric cabs, exhibiting a Rosenblatt-like heavy-tailed structure. [PITH_FULL_IMAGE:figures/full_fig_p021_15.png]
Figure 16
Figure 16. Figure 16: Failure of Noise Approximation-Optimization Cascade [PITH_FULL_IMAGE:figures/full_fig_p027_16.png]
Figure 17
Figure 17. Figure 17: Failure of Noise Approximation-Optimization Cascade under Rosenblatt process [PITH_FULL_IMAGE:figures/full_fig_p028_17.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

46 extracted references · 43 canonical work pages

  1. [27]

    A stochastic calculus for rosenblatt processes,

    P. ˇCoupek, T. E. Duncan, and B. Pasik-Duncan, “A stochastic calculus for rosenblatt processes,” Stochastic Processes and their Applications, vol. 150, pp. 853–885, 2022

    work page 2022

  2. [43]

    An infinite time horizon linear-quadratic control problem with a rosenblatt process,

    P. ˇCoupek, T. E. Duncan, B. Maslowski, and B. Pasik-Duncan, “An infinite time horizon linear-quadratic control problem with a rosenblatt process,” in 2018 IEEE Conference on Decision and Control (CDC) , 2018, pp. 4973–4977

    work page 2018

  3. [44]

    A scalar linear-quadratic two player stochastic differential game with a rosenblatt process noise,

    T. E. Duncan and B. Pasik-Duncan, “A scalar linear-quadratic two player stochastic differential game with a rosenblatt process noise,” in 2020 59th IEEE Conference on Decision and Control (CDC) , 2020, pp. 343–347

    work page 2020

  4. [1]

    Ingenhousz, Vermischte Schriften physisch-medicinischen Inhalts

    J. Ingenhousz, Vermischte Schriften physisch-medicinischen Inhalts. A copie can be obtained from Nabu press, ISBN 978-1286537770, 1785

    work page

  5. [2]

    R. Brown, “A brief account of microscopical observations made on the particles contained in the pollen of plants; and on the general existence of active molecules in organic and inorganic bodies,” Philosophical Magazine, vol. 4, no. 21, pp. 161–173, 1828

    work page

  6. [3]

    ¨Uber die von der molekularkinetischen theorie der w¨arme geforderte bewegung von in ruhenden fl¨ussigkeiten suspendierten teilchen,

    A. Einstein, “ ¨Uber die von der molekularkinetischen theorie der w¨arme geforderte bewegung von in ruhenden fl¨ussigkeiten suspendierten teilchen,” Annalen der Physik , vol. 322, no. 8, pp. 549–560, 1908

    work page 1908

  7. [4]

    B. B. Mandelbrot, The fractal geometry of nature . W.H. Freeman and company, 1982. 45

    work page 1982

  8. [5]

    Long-term storage capacity of reservoirs,

    H. E. Hurst, “Long-term storage capacity of reservoirs,” Transactions of the American Society of Civil Engineers , vol. 116, no. 1, pp. 770–808, 1951

    work page 1951

Show all 46 references
  1. [6]

    Cooperative mean-field type games,

    A. K. Cisse and H. Tembine, “Cooperative mean-field type games,” in 19th World Congress of the International Federation of Automatic Control (IFAC), Cape Town, South Africa, August 2014, pp. 8995–9000

    2014

  2. [7]

    Risk-sensitive mean-field-type games with lp-norm drifts,

    H. Tembine, “Risk-sensitive mean-field-type games with lp-norm drifts,” Automatica, vol. 59, pp. 224–237, 2015

    2015

  3. [8]

    Mean-field-type games,

    ——, “Mean-field-type games,” AIMS Mathematics, vol. 2, no. 4, pp. 706–735, 2017

    2017

  4. [9]

    Mean-field-type games driven by mixture of noises,

    T. E. Duncan, B. Pasik-Duncan, and H. Tembine, “Mean-field-type games driven by mixture of noises,” International Journal of Dynamics and Control, accepted and to appear , 2025

    2025

  5. [10]

    Bas ¸ar, B

    T. Bas ¸ar, B. Djehiche, and H. Tembine, Mean-Field-Type Game Theory: Foundations and New Directions , 2025, vol. 1, forthcoming

    2025

  6. [11]

    2, forthcoming

    ——, Mean-Field-Type Game Theory: Applications , 2025, vol. 2, forthcoming

    2025

  7. [12]

    Independence and dependence,

    M. Rosenblatt, “Independence and dependence,” in Proceedings of the Fourth Berkeley Symposium on Mathematical Statistics and Probability, Volume 2: Contributions to the Theory of Statistics . University of California Press, 1961, pp. 431–443

    1961

  8. [13]

    Central limit theorem for stationary processes,

    ——, “Central limit theorem for stationary processes,” pp. 551–561, 1972. [On- line]. Available: https://projecteuclid.org/proceedings/berkeley-symposium-on-mathematical-statistics-and-probability/ Proceedings-of-the-Sixth-Berkeley-Symposium-on-Mathematical-Statistics-and/Chapt...

    1972

  9. [14]

    Graduate Texts in Mathematics

    ——, Random Processes, ser. Graduate Texts in Mathematics. New York: Springer, 1974, vol. 17

    1974

  10. [15]

    Boston: Birkh ¨auser, 1985

    ——, Stationary Sequences and Random Fields . Boston: Birkh ¨auser, 1985

    1985

  11. [16]

    Weak convergence to fractional brownian motion and to the rosenblatt process,

    M. S. Taqqu, “Weak convergence to fractional brownian motion and to the rosenblatt process,” Advances in Applied Probability, vol. 7, no. 2, pp. 249–269, June 1975

    1975

  12. [17]

    Non-central limit theorems for non-linear functionals of gaussian fields,

    R. Dobrushin and P. Major, “Non-central limit theorems for non-linear functionals of gaussian fields,” Zeitschrift f¨ur Wahrscheinlichkeit- stheorie und Verwandte Gebiete , vol. 50, no. 1, pp. 27–52, 1979

    1979

  13. [18]

    Convergence of integrated processes of arbitrary hermite rank,

    M. S. Taqqu, “Convergence of integrated processes of arbitrary hermite rank,” Zeitschrift f¨ur Wahrscheinlichkeitstheorie und Verwandte Gebiete, vol. 50, no. 1, pp. 53–83, 1979

    1979

  14. [19]

    Multiple stochastic integrals with dependent integrators,

    R. Fox and M. S. Taqqu, “Multiple stochastic integrals with dependent integrators,” Journal of Multivariate Analysis , vol. 21, no. 1, pp. 105–127, 1987

    1987

  15. [20]

    Noncentral limit theorems and appell polynomials,

    F. Avram and M. S. Taqqu, “Noncentral limit theorems and appell polynomials,” Annals of Probability , vol. 15, no. 2, pp. 767–775, 1987

    1987

  16. [21]

    A bibliographical guide to self-similar processes and long-range dependence,

    M. S. Taqqu, “A bibliographical guide to self-similar processes and long-range dependence,” in Dependence in Probability and Statistics, ser. Progress in Probability and Statistics, E. Eberlein and M. Taqqu, Eds. Birkh ¨auser, 1986, vol. 11, pp. 137–162

    1986

  17. [22]

    Samorodnitsky and M

    G. Samorodnitsky and M. S. Taqqu, Stable Non-Gaussian Random Processes . London: Chapman and Hall, 1994

    1994

  18. [23]

    Non-gaussian properties of the real industrial control error in siso loops,

    P. D. Doma ´nski, “Non-gaussian properties of the real industrial control error in siso loops,” in 2015 19th International Conference on System Theory, Control and Computing (ICSTCC) , 2015

    2015

  19. [24]

    Solving groundwater depletion in india while achieving food security,

    N. Devineni, S. Perveen, and U. Lall, “Solving groundwater depletion in india while achieving food security,” Nature Communications, vol. 13, p. 3374, 2022

    2022

  20. [25]

    Adaptive multichannel combining and equalization for underwater acoustic communications,

    M. Stojanovic, J. Catipovic, and J. G. Proakis, “Adaptive multichannel combining and equalization for underwater acoustic communications,” Journal of the Acoustical Society of America , vol. 94, no. 3, pp. 1621–1631, 1993

    1993

  21. [26]

    Basin-scale acoustic communication: A feasibility study using tomography m-sequences,

    L. Freitag and M. Stojanovic, “Basin-scale acoustic communication: A feasibility study using tomography m-sequences,” in Proceedings of MTS/IEEE OCEANS , Honolulu, HI, USA, Nov. 2001, pp. 2256–2261

    2001

  22. [28]

    Besov-orlicz path regularity of non-gaussian processes,

    P. ˇCoupek and M. Ondrej ´at, “Besov-orlicz path regularity of non-gaussian processes,” Potential Analysis, vol. 60, no. 2, pp. 307–339, 2024

    2024

  23. [29]

    Young differential equations driven by besov-orlicz paths,

    F. ˇCoupek, J. Hendrych, and Slav ´ık, “Young differential equations driven by besov-orlicz paths,” Discrete & Continuous Dynamical Systems - B , 2025

    2025

  24. [30]

    Stochastic integration with respect to fractional processes in banach spaces,

    P. ˇCoupek, B. Maslowski, and M. Ondrej ´at, “Stochastic integration with respect to fractional processes in banach spaces,” Journal of Functional Analysis, vol. 282, no. 2, p. art. no. 109393, 2022

    2022

  25. [31]

    Parameter estimation and singularity of laws on the path space for sdes driven by rosenblatt processes,

    P. ˇCoupek, P. Kˇr´ıˇz, and B. Maslowski, “Parameter estimation and singularity of laws on the path space for sdes driven by rosenblatt processes,” Stochastic Processes and their Applications , vol. 179, p. art. no. 104499, 2025

    2025

  26. [32]

    Rough differential equations driven by besov-orlicz paths,

    P. ˇCoupek, J. Hendrych, and Slav ´ık, “Rough differential equations driven by besov-orlicz paths,” 2025

    2025

  27. [33]

    The need for non-gaussian noise in control system models. why non-gaussian noise matters?

    P. D. Doma ´nski, T. Duncan, and B. Pasik-Duncan, “The need for non-gaussian noise in control system models. why non-gaussian noise matters?” 2025, submitted

    2025

  28. [34]

    Stochastic control systems with long range dependent noise,

    T. E. Duncan and B. Pasik-Duncan, “Stochastic control systems with long range dependent noise,” in Outliers in Control Engineer- ing–Fractional Calculus Perspective, P. D. Doma ´nski, Y . C. Chen, and M. Lawrynczuk, Eds. Berlin Boston: De Gruyter, 2021, pp. 51–63

    2021

  29. [35]

    Advances in noise modeling for stochastic systems in optimal and adaptive control,

    ——, “Advances in noise modeling for stochastic systems in optimal and adaptive control,” LaMatematica (Official journal of the Assn. for Women in Math.) , inaugural edition

  30. [36]

    On a generalization of a stochastic integral,

    A. V . Skorokhod, “On a generalization of a stochastic integral,” Theory of Probability & Its Applications , vol. 20, no. 2, pp. 219–233, 1975

    1975

  31. [37]

    From forward integrals to wick-it ˆo integrals: the fractional brownian motion and the rosenblatt process cases,

    B. Arras, “From forward integrals to wick-it ˆo integrals: the fractional brownian motion and the rosenblatt process cases,” 2016, preprint

    2016

  32. [38]

    A white noise approach to stochastic integration with respect to the rosenblatt process,

    ——, “A white noise approach to stochastic integration with respect to the rosenblatt process,” Potential Analysis, vol. 43, no. 3, pp. 547–591, 2015

    2015

  33. [39]

    Multiple wiener integrals,

    K. It ˆo, “Multiple wiener integrals,” Journal of the Mathematical Society of Japan , vol. 3, no. 3-4, pp. 157–169, 1951

    1951

  34. [40]

    Ergodicity of increments of a rosenblatt process and some consequences,

    P. ˇCoupek, P. K ˇr´ıˇz, and B. Maslowski, “Ergodicity of increments of a rosenblatt process and some consequences,” Czech. Math. J. , vol. 75, no. 1, pp. 327–343, 2025. 46

    2025

  35. [41]

    Prediction and related topics for a scalar linear stochastic equation with a rosenblatt process noise,

    T. E. Duncan and B. Pasik-Duncan, “Prediction and related topics for a scalar linear stochastic equation with a rosenblatt process noise,” in 10th International Conference on Control, Decision and Information Technologies (CoDIT) , July 2024, pp. 1475–1478

    2024

  36. [42]

    Linear prediction for some rosenblatt and rosenblatt-volterra processes,

    ——, “Linear prediction for some rosenblatt and rosenblatt-volterra processes,” Journal of Systems Science and Complexity , vol. 38, no. 1, pp. 1–7, 2025

    2025

  37. [45]

    Mean-field-type games driven by rosenblatt processes,

    T. E. Duncan, B. Pasik-Duncan, and H. Tembine, “Mean-field-type games driven by rosenblatt processes,” in CoDIT 2024, 2024, pp. 982–987

    2024

  38. [46]

    Multi-layer mean-field-type games driven by five noises,

    ——, “Multi-layer mean-field-type games driven by five noises,” in IEEE CDC 2024 , 2024. Tyrone E. Duncan (M’92-SM’96-F’99) received his bachelor’s degree from Rensselaer Polytechnic Institute in 1963 and his master’s and doctoral degrees from Stanford University in 1964 and 19...

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.