Pith. sign in

REVIEW 3 major objections 4 minor 21 references

Robust Control under Stationary Ambiguity

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper shows that simulator policies lose robustness as parameter ambiguity vanishes, and that training under stationary ambiguity—a stationary filter over the latent state—preserves robustness across regime shifts.

desk verdict A genuinely useful practical idea (refresh latent model for robustness to shifting regimes) paired with a formal definition that is weaker than the paper's own claims; worth a serious referee because the experiments and the explicit limitations are honest. read the letter →

arxiv 2608.04832 v1 pith:JU7FXNIE submitted 2026-08-05 cs.LG q-fin.CP

Konrad J. Mueller , Amira Akkari , Ben Wood , Lukas Gonon This is my paper
classification cs.LGq-fin.CP MSC 93E2091G2060G35
keywords stationaryambiguitydomainrandomizationdeephedgingstochasticfilteringregimeshiftslatentparameteruncertaintyrefreshmodelsim-to-realrobustness
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

Simulator-based control policies often become overconfident when the simulator's unobserved parameter is drawn once and then fixed: the policy gradually infers the parameter, its ambiguity decays, and it specializes to one regime. This paper argues that in non-episodic systems, such as financial markets, regimes shift and this specialization is dangerous. The authors propose stationary ambiguity—requiring the simulator's filter process over the latent state to be stationary—so that uncertainty does not systematically decay with the control clock. A practical way to induce it is the refresh latent model, in which the latent parameter occasionally refreshes from a fixed distribution; policies trained this way stay robust to volatility and correlation shifts and outperform static randomization in an S&P 100 hedging backtest.

What carries the argument

The central object is the filter process π_t = P(X_t ∈ · | G_t), the conditional law of the latent simulator state given all observations, which the paper names the carrier of ambiguity; stationary ambiguity means this process is strictly stationary. The key result is Proposition 2: if the joint process (X, Y) is stationary on the bi-infinite time axis, the filter is stationary, which lets one randomize a base simulator by replacing the fixed parameter x with a stationary Markov process X. For Markov base simulators a contraction condition (Proposition 3, adapted from Stenflo) guarantees the joint process is stationary, and the refresh latent model X_t | X_{t-1} ∼ (1−α)δ_{X_{t-1}} + αν realizes this with a single extra parameter; a finite warm-up from the stationary distribution places the filter near its stationary regime at the start of control.

What would settle it

Take a simulator that satisfies the paper's stationary-ambiguity definition but has zero ambiguity, for instance a static latent model run on the bi-infinite time axis so the filter is constant and degenerate, train a policy under it, and subject it to the paper's regime-shift stress test: if it remains robust, stationarity alone suffices; the paper's own RLM experiments predict it will not.

Watch

Extended reading notes

Core claim

Under static randomization, where the simulator parameter is drawn once per trajectory, the filter concentrates over time (Doob's theorem) and the policy progressively specializes to its estimate of the parameter. The paper's central claim is that this vanishing ambiguity is the wrong inductive bias when latent parameters can shift, and that training in a simulator whose filter π_t = P(X_t ∈ · | G_t) is strictly stationary—called stationary ambiguity—yields policies that maintain continual robustness. The paper proves that stationary joint dynamics imply a stationary filter, constructs stationary randomization schemes (including the refresh latent model), and demonstrates on three hedging problems and a real-market backtest that these policies outperform static-randomization policies in regime-shift stress tests and on path-dependent payoffs.

Load-bearing premise

The key premise is that keeping the policy's uncertainty about the hidden regime stationary in time is what preserves robustness; however, a simulator can satisfy that condition while having zero uncertainty, so a positive and varying ambiguity level is the real requirement.

Editorial extensions

If this is right

  • Policies trained under the refresh latent model (RLM) maintain low hedging risk when the volatility or correlation regime changes mid-horizon, while static-latent (SLM) policies specialize to the pre-shift regime and degrade.
  • On an out-of-sample S&P 100 backtest across ten payoffs, RLM-trained policies achieve the lowest spectral risk on the harder payoffs and on the regime-shift subsample; SLM has only a slight edge on the four simple payoffs.
  • Simulator initialization matters: deterministically initializing latent states creates a systematic increase in ambiguity and violates stationary ambiguity; warm-up from the stationary distribution should be used instead.
  • The RLM's refresh probability α is a practical tuning knob: small α preserves base-simulator paths, large α drifts toward i.i.d. randomization and hurts fixed-regime performance; intermediate values give regime-shift robustness at little initial cost.
  • Aggregate in-simulator evaluations can miss robustness defects; controlled regime-shift stress tests are needed.

Reading between the lines

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

  • The paper's formal definition of stationary ambiguity is under-specified: a static latent model on the bi-infinite time axis has a stationary filter with zero ambiguity, so any practical deployment should add an explicit requirement of positive, state-dependent ambiguity (e.g., ergodicity), not just stationarity.
  • The same probe methodology used to show that policies 'learn to filter' could be turned into an audit tool for continual robustness: measure how quickly a policy's internal uncertainty band widens after a suspected regime shift, without needing a full stress test.
  • Because the RLM forgets the remote past exponentially, it behaves like an adaptive constant-gain learner; a testable extension is to select the refresh probability α from the time-scale of regime shifts in a domain, and to compare RLM-trained policies with explicit constant-gain belief-updating rules.
  • The principle is restricted to control problems where actions do not affect the observation process; in exploration problems, stationary ambiguity would suppress information-gathering, so the boundary between persistent-ambiguity settings and exploration settings is itself a useful contribution.
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 proposes a simulator-design principle called 'stationary ambiguity' for sequential control problems driven by exogenous processes. The authors argue that commonly used static randomization of a latent parameter induces a filter over that parameter that concentrates over time, so the trained policy specializes and loses robustness to later regime changes. They formalize the desired property as strict stationarity of the filter process, prove that stationarity of the joint latent-observation process is sufficient, and introduce the refresh latent model (RLM) as a practical way to obtain it. They illustrate the distinction in a closed-form optimal-investment example and in three neural hedging problems, and they report a real-market backtest on S&P 100 stocks in which RLM-trained policies outperform static-latent-model policies on path-dependent payoffs and in high-regime-shift periods.

Significance. If the formal definition is tightened, this is a valuable and generally applicable contribution. It identifies a real sim-to-real mismatch in randomized simulator training, gives a simple one-parameter construction (RLM) that preserves the base simulator between refreshes, and supports the claims with both clean analytical examples and extensive experiments. The closed-form investment analysis (Propositions 4-6), the filter-stationarity argument (Proposition 2), and the out-of-sample protocol (prior fitted on 2006-2015 data, backtested on 2016-2025 data) are strengths. The evaluation is not circular: the RLM is not fitted to the test-period performance, and the comparison is against external Black-Scholes baselines.

major comments (3)
  1. [Section 3.2, Definition and 'Stronger ambiguity conditions'] The formal definition of stationary ambiguity is too weak to carry the central claim that policies trained under it maintain continual robustness. On the bi-infinite time axis, the static latent model with X_t ≡ X_1 is a stationary joint process, so Proposition 2 applies; by Proposition 1, the filter is π_t = δ_{X_1} with V_t^φ = 0 almost surely. That simulator therefore satisfies the stated definition while inducing zero ambiguity, and it is indistinguishable from a plug-in simulator in terms of robustness. The pathwise-constant example V_t^φ = v_low 1_{B=0} + v_high 1_{B=1} also passes the definition while never mixing across paths. The paper acknowledges this in the 'Stronger ambiguity conditions' paragraph, but the abstract and contributions still present stationary ambiguity as 'the modeling principle' that guarantees continual robustness, and the experiments attribute their success to it. The positive, refreshing, ergodic ambiguity of the RLM is an additional assumption that stationarity alone does not imply. Please either strengthen the definition (for example, require the filter process to be ergodic and to have a non-degenerate stationary distribution) or explicitly reframe stationary ambiguity as a necessary condition, and adjust the abstract and contribution claims accordingly.
  2. [Section 3.4, finite-warm-up scheme] The warm-up construction only provides an asymptotic guarantee in H, and the paper states this explicitly: 'for a chosen warm-up length H < ∞, this asymptotic statement does not quantify the difference in ambiguity between the finite-warm-up and infinite-past simulators.' This matters for the practical recipe in Section 6.2, which recommends warm-up to place the filter 'approximately' in its stationary regime. For the RLM with α = 0.01, the expected refresh interval is 100 time steps, while the real-data experiments use H = 32 and the synthetic experiments use H = 64 or longer; the warm-up period is shorter than the expected refresh time. A quantitative filter-forgetting bound for the specific models, or at least a sensitivity analysis showing that the main conclusions are stable across H and α, is needed to support the claim that the finite-warm-up policy is operating under approximately stationary ambiguity.
  3. [Section 5.2 and Fig. 10] The real-data comparison reports pooled spectral risk losses without uncertainty quantification. The losses are pooled over overlapping 128-day windows and across many stocks, so the observations are strongly dependent; the plotted 'improvement over baseline' values have no confidence intervals or paired tests. This makes it hard to assess whether the RLM advantage on the six path-dependent payoffs is statistically meaningful, especially because the four simple payoffs show no clear difference. Please add standard errors or block-bootstrap confidence intervals for the pooled risk estimates, or at least for the headline comparisons between SLM and RLM.
minor comments (4)
  1. [Appendix C.2.3, Proposition 6] The formula for Var(ε_t) requires the constraint φ^2 < σ^2/(σ^2 − τ^2) for positivity; the paper does not state this constraint, although the chosen numerical parameters satisfy it.
  2. [Section 5.1] The data description says the universe is stocks that made up the S&P 100 at the end of December 2015, but it does not report how many stocks remain in the sample or how the index membership changes are handled over 2016-2025.
  3. [Table 2] The column headings 'SHIFT MAX GAP' and 'SHIFT MIN GAP' are unclear on first reading; the caption defines them only through the footnote, and the table would benefit from a direct statement that these are the regime pairs where the SLM-RLM difference is largest or smallest.
  4. [Section 6.2] The 'conditional initialization' scheme requires sampling from the filter P(X_0 ∈ · | Y_{1-H:0} = y_{1-H:0}), but no numerical method is suggested for this step; since the warm-up schemes are the ones actually used, this is a presentation issue rather than a blocking one.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's empirical comparisons are against external benchmarks and its formal claims are proven from stated assumptions rather than imported from self-citation.

full rationale

The paper's derivation chain is self-contained and does not reduce to its own inputs by construction. The main formal results are Proposition 1 (Doob's posterior consistency theorem), Proposition 2 (sufficient condition for filter stationarity), and Proposition 3 (Stenflo's contraction condition for stationary random-parameter Markov simulators). All three are either standard external theorems or are proven in Appendix C with explicit assumptions stated in the main text; none is a uniqueness theorem imported from the authors' prior work. The experiments compare SLM and RLM policies trained under the same randomization distribution ν, where ν is fit to pre-2016 historical data and the backtest is on out-of-sample 2016-2025 data, so the predictive claims are not forced by fitted parameters. The refresh probability α is chosen by design considerations and is not tuned to the test outcomes. The only soft spot is that the formal definition of stationary ambiguity is weaker than the intended notion, since it admits zero-ambiguity stationary filters and non-ergodic ambiguity patterns; however, the paper explicitly acknowledges this in Section 3.2 ('Filter stationarity should not be understood as a sufficient condition for representing ambiguity realistically') and the experimental benefit is attributed to the disclosed RLM construction, whose positive, refreshing, ergodic dynamics are exactly what the simulations use. This is an under-specification of the modeling principle, not a circular derivation. Self-citations (e.g., Mueller et al. 2026) appear only as illustrative examples and related work, not as load-bearing justification for the central claims. Therefore no circular step is present.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

No new physical entities are introduced. The refresh latent model is a stochastic process construction, not an invented entity such as a new particle or force, and it is validated through experiments rather than posited as an unobserved cause.

free parameters (5)
  • RLM refresh probability alpha = 0.01
    Hand-chosen; controls how often the latent parameter refreshes. Section 4.1 and 5.1. Fig. 9 shows sensitivity.
  • Prior volatility distribution nu for real-data study = InvGamma(alpha_IG=1.63, beta_IG=0.07)
    Maximum-likelihood fit to 128-day return windows from 2006-2015 S&P 100 stocks; used to sample latent volatility in SLM/RLM simulators for the backtest.
  • Synthetic experiment priors = InvGamma(5.93,0.16), Beta(1.80,5.50), Beta(4.00,2.00)
    Hand-chosen for Section 4 hedging problems; define the ambiguity range and the regime-shift stress tests.
  • Volatility cap in training = 1.0
    Appendix B.1.3: sampled volatility values are capped at 1.0 during training to avoid excessive sensitivity; a post hoc adjustment of the randomization distribution.
  • Warm-up length H = 64 in Section 4, 32 in Section 5
    Finite substitute for the infinite-past filter; no quantitative convergence check that the filter is near stationarity at this H.
assumptions (4)
  • domain assumption The real market's latent volatility/correlation is well modeled by a stationary Markov process (RLM) with fixed prior nu and constant alpha.
    Used in Section 5 backtest; if the real regime-switching rate or prior is misspecified, stationary ambiguity may be the wrong model.
  • domain assumption Finite warm-up H is long enough to approximate the stationary filter.
    Section 6.2 warm-up scheme; the paper gives only asymptotic convergence as H goes to infinity and no rate for the models used.
  • domain assumption Parameter uncertainty is the only relevant sim-to-real gap; base simulator dynamics are otherwise correct.
    Section 2.1: plug-in and randomized simulators share the same base model F; model misspecification beyond parameters is not addressed.
  • domain assumption Neural policies are expressive enough to approximate the optimal filter-based controls.
    Section 4.2 probes show partial evidence, but the policy class is not proven sufficient.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust Control under Stationary Ambiguity." pith.science (2026). https://pith.science/paper/JU7FXNIE

@misc{pith2026260804832,
  author       = {Pith},
  title        = {Pith review of: Robust Control under Stationary Ambiguity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JU7FXNIE}},
  note         = {Machine review of arXiv:2608.04832}
}
abstract

Control policies optimized in simulation can perform poorly in the real system when the parameters $x$ of the simulator are estimated from limited data but the resulting parameter uncertainty is not represented inside the simulation. A common way to incorporate such ambiguity is to simulate each trajectory of the system under a randomly drawn value for $x$. Since the policy cannot observe the drawn value, it must initially choose controls that perform well across many possible parameter values. However, if the policy progressively observes the system, it can often gradually infer the value of $x$, so that ambiguity vanishes. Over time, the policy then specializes to its estimate of $x$ and loses its robustness. This is undesirable in many real systems, where latent factors are expected to shift. In financial markets, for example, a policy hedging a derivative payoff should remain robust to changes in the volatility regime. To induce such continual robustness, we propose training policies in simulators where ambiguity varies with the system's state but does not systematically decay over time. We formalize this requirement as stationary ambiguity: the simulator should induce a stationary filter process over the latent state. We show how to construct such simulators and demonstrate, on hedging problems, that policies trained under stationary ambiguity preserve robustness to latent factors over time, leading to strong performance on real market data. As a modeling principle, stationary ambiguity informs many simulator design decisions: which models make realistic simulators, how their parameters should be randomized, and how simulator and policy should be initialized. While our experiments focus on hedging, stationary ambiguity may also be useful for other sequential control problems driven by exogenous stochastic processes with shifting latent structure.

Figures

Figures reproduced from arXiv: 2608.04832 by the authors.

Figure 1
Figure 1. Ambiguity decay leads to robustness decay in a [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Vanishing ambiguity leads to specialization of optimal investment. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Static randomization policy fails to adapt to a regime switch. We evaluate the policies on scenarios where the [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (15 more)
Figure 4
Figure 4. Figure 4: Neural hedging policies learn to filter. We probe the recurrent state [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 6
Figure 6. Figure 6: Continual robustness under volatility uncertainty ( [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 5
Figure 5. Figure 5: Initial robustness from simulator randomization. [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 8
Figure 8. Figure 8: Continual robustness under correlation uncertainty. [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 7
Figure 7. Figure 7: Continual robustness under covari￾ance regime shifts (BS-COV). We vary both σ1 and ϱ between the pre-trading and trad￾ing periods. Each cell groups scenarios by ϱ eval −ϱ pre and σ eval 1 −σ pre 1 , and shows the difference in semi-deviation loss between the policies (…
Figure 9
Figure 9. Figure 9: Sensitivity to the refresh probability in [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]
Figure 10
Figure 10. Figure 10: Hedging performance on S&P 100 backtest. Each point reports, for one payoff and one policy, [PITH_FULL_IMAGE:figures/full_fig_p017_10.png]
Figure 11
Figure 11. Figure 11: Minimax regret policy fails to adapt to a regime [PITH_FULL_IMAGE:figures/full_fig_p021_11.png]
Figure 12
Figure 12. Figure 12: Continual robustness results for BS-VOL. 31 [PITH_FULL_IMAGE:figures/full_fig_p031_12.png]
Figure 13
Figure 13. Figure 13: Continual robustness results for BS-VOL, when the policy is optimized with Rvar. A.2.3 BS-VOL (risk reversal) 0 15 30 45 Variance SLM (abs.) RLM (abs.) 0 15 30 45 SLM (rel.) RLM (rel.) 0.11 0.17 0.24 0.30 0.4 0.8 1.2 1.6 Semi-deviation 0.0 0.4 0.8 1.2 0.11 0.17 0.24 0…
Figure 14
Figure 14. Figure 14: Continual robustness results for BS-VOL when the payoff is a risk reversal. 32 [PITH_FULL_IMAGE:figures/full_fig_p032_14.png]
Figure 15
Figure 15. Figure 15: Continual robustness results for HESTON-CORR. A.2.5 BS-COV 4 8 12 Variance SLM (abs.) RLM (abs.) 0 4 8 12 SLM (rel.) RLM (rel.) -0.56 -0.06 0.44 0.93 0.2 0.4 0.6 0.8 Semi-deviation 0.0 0.2 0.4 0.6 -0.56 -0.06 0.44 0.93 1.5 3.0 4.5 CVaR 0.0 0.5 1.0 1.5 -0.56 -0.06 0.44…
Figure 16
Figure 16. Figure 16: Continual robustness results for BS-COV. 33 [PITH_FULL_IMAGE:figures/full_fig_p033_16.png]
Figure 17
Figure 17. Figure 17: Density of the randomization distribution [PITH_FULL_IMAGE:figures/full_fig_p035_17.png]
Figure 18
Figure 18. Figure 18: Additional results for the pooled S&P 100 real data backtest. Each point reports, for one payoff and one [PITH_FULL_IMAGE:figures/full_fig_p036_18.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

21 extracted references · 15 canonical work pages

  1. [1]

    Proposition(Doob’s theorem).Let X and Y be Polish spaces, equipped with their Borel σ-algebras

    This is a version of Doob’s well-known posterior consistency theorem; see, for example, Miller (2018). Proposition(Doob’s theorem).Let X and Y be Polish spaces, equipped with their Borel σ-algebras. Let ν be a probability measure on X , and let p(· |x)be a probability measure on Y for each x∈ X. Assume that x7→p(A|x) is measurable for every Borel setA⊆ Y....

    work page 2018

  2. [2]

    The result is well-known (Kunita, 1971; van Handel, 2012)

    C.1.2 Filter stationarity We state a more general version of Proposition 2 and provide a proof for completeness. The result is well-known (Kunita, 1971; van Handel, 2012). A much more difficult question is under what conditions the filter has a unique stationary distribution (see e.g., van Handel, 2012). Proposition(Sufficient conditions for a stationary ...

    work page 1971

  3. [3]

    Our recursion is of this form, with the latent process X as the stationary driving sequence

    This result is due to Stenflo (2001), who studies iterations in which the map applied at each step is drawn along a stationary sequence. Our recursion is of this form, with the latent process X as the stationary driving sequence. Since the setup and notation in that work differ substantially from ours, we give the argument in detail. Recall that the rando...

    work page 2001

  4. [6]

    Since the drift is known, the policy does not learn about it from its observations and the optimal investment amount is constant over time

    C.2.1 Known drift We begin with the base simulator in which the drift is known and constant. Since the drift is known, the policy does not learn about it from its observations and the optimal investment amount is constant over time. Proposition 4(Known drift).Suppose that ¯x∈R, σ2 >0 , and Yt iid ∼N(¯x, σ2) for t= 1, . . . , T. Then the unique optimizer o...

    work page 2007

  5. [8]

    Exponential forgetting of smoothing distributions for pairwise Markov models

    Jüri Lember and Joonas Sova. Exponential forgetting of smoothing distributions for pairwise Markov models.ArXiv preprint arXiv:2103.05474,

    work page Pith review arXiv

  6. [13]

    Domain randomization for transferring deep neural networks from simulation to the real world

    Josh Tobin, Rachel Fong, Alex Ray, Jonas Schneider, Wojciech Zaremba, and Pieter Abbeel. Domain randomization for transferring deep neural networks from simulation to the real world. In2017 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2017, Vancouver, BC, Canada, September 24-28, 2017, pages 23–30. IEEE,

    work page 2017

  7. [14]

    Multi-Asset Spot and Option Market Simulation.ArXiv preprint arXiv:2112.06823,

    26 Magnus Wiese, Ben Wood, Alexandre Pachoud, Ralf Korn, Hans Buehler, Phillip Murray, and Lianjun Bai. Multi-Asset Spot and Option Market Simulation.ArXiv preprint arXiv:2112.06823,

    arXiv

  8. [20]

    Throughout this subsection, we work on the bi-infinite time axis and write Gt :=σ(Y s :s≤t), t∈Z

    This results in a scalar linear Gaussian state-space model (LG-SSM), so the filtering equations are given by the Kalman filter (see e.g., Durbin and Koopman, 2012). Throughout this subsection, we work on the bi-infinite time axis and write Gt :=σ(Y s :s≤t), t∈Z. We consider the stationary version of the model, so that the Kalman filter is in steady state ...

    work page 2012

Show all 21 references
  1. [21]

    It therefore does not force the policy to inherit the specialization behavior of the static latent policy

    This parameterization lets the policy use both the full historical sample mean ¯Yt−1 and the Kalman mean mKF t , which forgets older observations. It therefore does not force the policy to inherit the specialization behavior of the static latent policy. The functionsw1,θ, w2,θ...

    1999

  2. [1993]

    Designing and Interpreting Probes with Control Tasks.ArXiv preprint arXiv:1909.03368,

    John Hewitt and Percy Liang. Designing and Interpreting Probes with Control Tasks.ArXiv preprint arXiv:1909.03368,

    1909 arXiv

  3. [1995]

    Jeffrey W. Miller. A detailed treatment of Doob’s theorem.ArXiv preprint arXiv:1801.03122,

    arXiv

  4. [2001]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A method for stochastic optimization.ArXiv preprint arXiv:1412.6980,

    arXiv

  5. [2008]

    Pérez-Mendoza

    Pascal François, Geneviève Gauthier, Frédéric Godin, and Carlos O. Pérez-Mendoza. Deep Hedging with Options Using the Implied V olatility Surface.ArXiv preprint arXiv:2504.06208,

    arXiv

  6. [2009]

    Exponential Spectral Risk Measures.ArXiv preprint arXiv:1103.5409,

    Kevin Dowd and John Cotter. Exponential Spectral Risk Measures.ArXiv preprint arXiv:1103.5409,

    arXiv

  7. [2013]

    Deep Learning Approximation for Stochastic Control Problems.ArXiv preprint arXiv:1611.07422,

    Jiequn Han and Weinan E. Deep Learning Approximation for Stochastic Control Problems.ArXiv preprint arXiv:1611.07422,

    arXiv

  8. [2014]

    Sim-to-Real Transfer of Robotic Control with Dynamics Randomization

    Xue Bin Peng, Marcin Andrychowicz, Wojciech Zaremba, and Pieter Abbeel. Sim-to-Real Transfer of Robotic Control with Dynamics Randomization. In2018 IEEE International Conference on Robotics and Automation, ICRA 2018, Brisbane, Australia, May 21-25, 2018, pages 1–8. IEEE,

    2018

  9. [2019]

    Lyons, Imanol Pérez Arribas, and Ben Wood

    Hans Buehler, Blanka Horvath, Terry J. Lyons, Imanol Pérez Arribas, and Ben Wood. A Data-driven Market Simulator for Small Data Environments.ArXiv preprint arXiv:2006.14498,

    2006 arXiv

  10. [2020]

    Uncertainty-Aware Strategies: A Model-Agnostic Framework for Robust Financial Optimization through Subsampling.ArXiv preprint arXiv:2506.07299,

    Hans Buehler, Blanka Horvath, Yannick Limmer, and Thorsten Schmidt. Uncertainty-Aware Strategies: A Model-Agnostic Framework for Robust Financial Optimization through Subsampling.ArXiv preprint arXiv:2506.07299,

    arXiv

  11. [2022]

    Option Valuation and Hedging Strategies with Jumps in the V olatility of Asset Returns.The Journal of Finance, 48 (5):1969–1984,

    Vasanttilak Naik. Option Valuation and Hedging Strategies with Jumps in the V olatility of Asset Returns.The Journal of Finance, 48 (5):1969–1984,

    1969

  12. [2023]

    Parametric regret in uncertain Markov decision processes

    Huan Xu and Shie Mannor. Parametric regret in uncertain Markov decision processes. InProceedings of the 48th IEEE Conference on Decision and Control, CDC 2009, Combined with the 28th Chinese Control Conference, December 16-18, 2009, Shanghai, China, pages 3606–3613. IEEE,

    2009

  13. [2024]

    Mueller, Nikita Zozoulenko, Ben Wood, Thomas Cass, and Lukas Gonon

    Konrad J. Mueller, Nikita Zozoulenko, Ben Wood, Thomas Cass, and Lukas Gonon. Generating Financial Time Series by Matching Random Convolutional Features.ArXiv preprint arXiv:2606.05138,

    arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.