Pith. sign in

REVIEW 2 major objections 4 minor 47 references

Tamed Euler Schemes for Singular SDEs with Multiplicative Levy Noise

T0 review · 2 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read Tamed Euler schemes get explicit strong convergence rates for singular SDEs with multiplicative Lévy noise.

desk verdict Genuinely new error decomposition, but the main theorem drops a β-dependent term that the paper's own Proposition 8.8 says to keep — the stated rate isn't proven for all allowed parameters. read the letter →

arxiv 2607.18862 v1 pith:JXA3S4JV submitted 2026-07-21 math.PR

Hua Zhang , Mingbo Zhang This is my paper
classification math.PR MSC 60H3560H1060H5060L9035B65
keywords tamedEuler–MaruyamasingulardriftLévynoisestrongconvergencerateZvonkintransformKrylovestimatesstochasticGronwalljumpdiscretizationerror
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

This paper aims to show that an explicit tamed Euler–Maruyama scheme converges to the solution of a multidimensional SDE whose drift is only Ladyzhenskaya–Prodi–Serrin integrable, whose Brownian and jump coefficients have Sobolev regularity, and whose jumps are multiplicative. The theorem gives a strong-error bound with a rate made of five additive pieces: the initial error, the drift-approximation error, the Brownian discretization error, and two genuinely jump-induced errors from freezing the jump coefficient and the nonlocal compensator on the time grid. A sympathetic reader should care because previously explicit rates for such irregular jump SDEs were not available; if the theorem holds, numerical simulations of these equations can be run with a quantitative guarantee. When the jump coefficient vanishes, the two jump terms disappear and the estimate reduces to the Brownian-type result.

What carries the argument

The load-bearing machinery is a nonlocal Zvonkin transform: one solves a backward parabolic integro-differential equation whose solution u has the same drift term as the SDE, so that applying Itô's formula to u(t,X_t) removes the singular drift from the error identity. Around this, the paper builds four supporting estimates: a discrete Krylov estimate for the Euler path (L^q_p functionals along X^n are bounded uniformly in n), parabolic a priori estimates for the Zvonkin equation with distributional forcing, pathwise Burkholder–Davis–Gundy inequalities for compensated Poisson integrals, and a stochastic Davie–Gronwall lemma that closes the moment bound. The main structural device is a ten-te

What would settle it

Run a one-dimensional test case with singular drift b(x)=|x|^{-γ} (truncated at scale 1/n), σ≡1, and a multiplicative compound Poisson jump term g(x,z)=c·x·z·1_{|z|<1}. For p̄≥2, the jump-induced part of the error should scale like n^{−(1−1/q)(1−d/p)}; if the empirical strong L^{θ p̄} error changes at a different rate when c is varied, the theorem's predicted jump terms are not the correct ones.

Watch

Extended reading notes

Core claim

Under Conditions Hσ^g, Hb, and cHσ^g, the paper proves that for every θ∈(0,1) and admissible p̄, E[sup_{t∈[0,1]} |X^n_t−X_t|^{θ p̄}] ≤ N E_n^θ. For p̄≥2, E_n combines the initial error, the drift-approximation norm (ϖ_b^n(p̄))^{p̄}, Brownian terms n^{−α/2}, n^{−1/2} log n and n^{−α(1−1/q)}, and jump terms C_J^{(m)} n^{−(1−1/q)} and C_J^{(c)} n^{−(1−1/q)(1−d/p)}; analogous powers hold for 1<p̄<2. This is, to the authors' knowledge, the first explicit strong convergence rate for tamed Euler schemes for LPS-singular drift with multiplicative Lévy noise. The rate is not meant to compete with the classical order 1/2 for Lipschitz coefficients; it quantifies the price of allowing singular drift an

Load-bearing premise

The load-bearing premise is Condition (Hb)'s scale control: every tamed drift b^n must lie in L^q_p∩L^q_∞ and satisfy (1/n)^{1/2−1/q}‖b^n‖_{L^q_∞([s,t])} ≤ m_n(s,t)^θ with sup_n m_n(0,1)<∞; if a natural regularization such as b^n=b with only LPS integrability does not meet this L^q_∞ scale condition, the Girsanov-weighted Krylov estimate and the Euler-increment bounds collapse, and the proof of the convergence theorem does not close.

Editorial extensions

If this is right

  • Numerical simulation of singular jump SDEs can be justified with explicit strong-error rates: for a given mesh n, the error bound grows only through the five listed terms, so users know which part of the rate is limiting.
  • Drift approximation and stochastic discretization are cleanly separated in the bound, so one can choose mollified, truncated, wavelet, or φ-transform approximations and obtain concrete rates once their negative-order error is controlled.
  • If the jump coefficient is constant or spatially flat, the two jump-induced terms vanish and the estimate recovers the Brownian-only rate; this makes the jump contribution visible as an explicit extra cost of multiplicative jumps.
  • For the small-jump setting treated in the paper, the rates show the jump-induced loss is n^{−(1−1/q)} (compensated martingale) and n^{−(1−1/q)(1−d/p)} (nonlocal compensator); these are the terms to watch in any extension to larger jumps or different Lévy measures.

Reading between the lines

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

  • An implication left implicit: the error decomposition isolates which noise source degrades the rate, so in applications one could design the mesh to balance the jump-induced terms against the Brownian terms—for instance, raising q reduces n^{−(1−1/q)} but leaves the d/p dependence in the nonlocal term.
  • A testable extension is to use the same W8/W9 decomposition for SDEs with large jumps (|z|≥R); the paper says its framework retains theoretical validity there, and a concrete rate would follow from the same two error mechanisms.
  • The omitted proof of Corollary 2.9 (rates for mollified, truncated, and wavelet drift approximations) is the missing link between the abstract ϖ_b^n bound and ready-to-run parameter choices; completing that argument would make the theorem directly usable in code.
  • The logarithmic factor from the stochastic Davie–Gronwall argument is plausibly removable; if so, the Brownian terms could be sharpened to match square-moment estimates available in the purely Brownian setting.
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

2 major / 4 minor

Summary. The paper develops a strong-convergence theory for tamed Euler–Maruyama schemes for multidimensional SDEs with LPS-type singular drift, multiplicative non-degenerate Brownian noise, and multiplicative Lévy noise. The central result, Theorem 2.4, gives an explicit strong rate in terms of an initial error, a pathwise drift-approximation error ϖ_b^n(\bar p), Brownian discretization errors, and two new jump-induced terms: a compensated-jump martingale freezing error and a nonlocal compensator freezing error. The proof is based on a nonlocal Zvonkin transform, discrete Krylov-type estimates, pathwise BDG inequalities, and a stochastic Gronwall argument. Theorem 2.7 provides bounds on the drift-approximation term, and Corollary 2.9 gives concrete rates for mollified and truncated drifts. The claimed novelty is the first explicit strong rate for this class of jump SDEs with singular drift, together with a clean separation of Brownian and jump error mechanisms.

Significance. If the main theorem is correct, this is a substantial contribution to numerical analysis of singular SDEs with jumps. The decomposition isolating the compensated-jump martingale error and the nonlocal compensator error is a genuine conceptual step beyond the Brownian theory of Lê–Ling, and the paper provides a large analytic apparatus (nonlocal Zvonkin PDE estimates, discrete Krylov estimates, stochastic Gronwall lemmas) that will likely be useful beyond the present setting. The paper honestly separates the deterministic drift-approximation error from stochastic discretization errors, and the use of explicit negative-Sobolev norms for the drift approximation is well motivated. However, two load-bearing gaps in the proof of Theorem 2.4 prevent the results from being accepted as they stand: a missing (1/n)^{1−β/2} term in the final rate, and an unjustified step in the derivation of the jump-specific rates. These are fixable in principle, but they require real mathematical work.

major comments (2)
  1. [§8.3, Eq. (8.19), and Theorem 2.4] Proposition 8.8 estimates the frozen-drift term W2 as containing (1/n)^{α/2} + (1/n)^{1−β/2} + (1/n)^{1/2}\log n, and then states that in the final theorem this is absorbed into the Brownian contribution 'under the standing restrictions on the parameters', with the caveat that otherwise (1/n)^{1−β/2} should be kept. However, the hypotheses of Theorem 2.4 (Conditions Hσ^g, Hb, cHσ^g) impose no relation between α and β that would make n^{−α/2} dominate n^{−(1−β/2)}. For example, d=3, α=0.8, β=1.5, q=10, p=100 satisfy the stated conditions; then (1/n)^{1−β/2} = n^{−0.25} is larger than n^{−α/2}=n^{−0.4}, n^{−1/2}\log n, n^{−α(1−1/q)}=n^{−0.72}, and the jump terms n^{−0.9}, n^{−0.873}. No term in E_n has size n^{−0.25}. Thus Theorem 2.4 is not proven as stated. The fix is either to add (1/n)^{1−β/2} to E_B,n or to impose a clear domination condition such as α+β<2 and verify it is compatible
  2. [§8.2, Proposition 8.7, Eqs. (8.15)–(8.18)] The derivation of the jump-specific rates is compressed to the point of being unverifiable. For W8, the argument moves from E[∫ |X^n_r−X^n_{r_n}|^{\bar p} ... dr] to an application of Lemma 8.5(8.9), which requires the exponent m>2. In the W9 estimate the exponent m = \bar p(1−d/p) can be less than 2 for \bar p∈[2, 2/(1−d/p)), and Theorem 2.4 does not exclude this. For W9, the sentence 'By Hölder's inequality in time and the Krylov-type estimate ... this yields' replaces an L^{\bar p}(Ω) norm of a time integral by ∫ E|X^n_r−X^n_{r_n}|^{\bar p(1−d/p)} dr; this move does not follow from standard Hölder/Jensen inequalities. Moreover, the Krylov-type estimates in Section 6 apply to integrands of the form h(r,X^n_r)(f(r,X^n_r)−f(r,X^n_{r_n})), not to powers of the Euler increment alone. The claimed exponents in (8.16) and (8.18) therefore lack a supporting proof. This is load-bearing because
minor comments (4)
  1. [Theorem 2.4] The condition defining the admissible range for \bar p is garbled in the text: '1, 2/d (p0∧p∧ p/(pd))' is not readable. The authors should restate this as a precise interval, since the allowable \bar p matters for the exponents in E_n.
  2. [Condition (Hb)] The notation 'p∈(2(1−α)^{-1}(d/β∨1)∨2,∞)' is ambiguous. Parentheses should be clarified or the condition split into explicit constraints on α, β, d, p.
  3. [Theorem 2.7, part (iii)] The control W0(s,t) in Theorem 2.7 conflicts notationally with the transformed-error term W0_t in Proposition 8.6. Rename one of them to avoid confusion.
  4. [Proposition 8.7] In the statement, the constants C_J^{(m)} and C_J^{(c)} are defined with powers \bar p, but the final estimates in (8.15)–(8.18) write them without powers on the right-hand side. The intended meaning is clear, but the notation should be made consistent with Theorem 2.4.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the rate is an honest two-level error decomposition, not a fitted parameter or self-citation loop.

full rationale

The paper's main bound (Theorem 2.4) expresses the strong error through E_n = E|x_0−x^n_0|^\bar p + (ϖ_b^n(\bar p))^\bar p plus explicit Brownian and jump discretization terms. The drift approximation term ϖ_b^n(\bar p) is defined in Definition 2.3 as the pathwise L^\bar p norm of ∫(I+∇U)[b−b^n](r,X_r)dr, i.e., an actual approximation error of b^n to b along the exact solution, and Theorem 2.7 independently estimates it by deterministic norms such as ||b−b^n||_{L^{q_1}_{p_1}} and ||b−b^n||_{H^q_{−θ,p}}. This is a genuine two-level structure: the Zvonkin/PDE estimates are auxiliary bounds, not a restatement of the target rate. The jump constants C_J^{(m)} and C_J^{(c)} are norms of the jump coefficient appearing in the assumptions and are then shown in Proposition 8.7 to control the two jump freezing errors; this is a forward estimate, not a parameter fitted to the conclusion. No uniqueness theorem or ansatz from the authors' own prior work is load-bearing; references such as [26] and [43] are external, and the arguments are either reproduced or cited as standard tools. The only defect visible in the text is a proof gap flagged by the authors themselves in Proposition 8.8: the W2 estimate contains a term (1/n)^{1−β/2} that Theorem 2.4's E_n does not carry, with the warning 'If one does not impose the corresponding domination condition, the term (1/n)^{1−β/2} should be kept explicitly in the final estimate.' That is a possible correctness/rate issue, not a circular input–output reduction, and per the review rules it does not increase the circularity score.

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

No data-fitted constants or invented physical entities appear. The theorem's many exponents (p,q,p0,q0,\bar p,θ,α,β) are regularity hypotheses, not fitted parameters. The main external inputs are the well-posedness and PDE regularity results of [43], plus standard Girsanov and stochastic-Gronwall tools.

assumptions (4)
  • domain assumption Strong well-posedness and Krylov-type estimate for the exact SDE (1.1) are imported from Xie–Zhang [43, Thm 2.2 and 5.6].
    Proposition 7.3 treats existence, uniqueness and E_s∫f(r,X_r)dr ≤ N∥f∥_{L^q_p} as imported facts. If these fail, Theorem 2.4 has no meaningful subject.
  • standard math Zvonkin backward PDE solvability and L^q_p regularity for the nonlocal integro-differential equation (Section 5) are established using the continuity method and results from [43, Thm 4.3].
    Lemma 8.1 and all uniform bounds for ∇u and ∇²u used in the error decomposition rely on this regularity theory.
  • standard math Girsanov theorem for Brownian motion plus Poisson random measure is used without proof in Theorems 4.7, 6.1 and 7.1.
    The Girsanov change of measure is a standard tool in this literature and is invoked repeatedly to reduce the tamed-drift case to the drift-free Euler scheme.
  • standard math Stochastic Davie–Gronwall lemma (Lemma 2.11) is imported from [12], and the quantitative Khasminskii lemma (Lemma 2.13) is proved in Appendix A using standard exponential-moment arguments.
    The final stochastic Gronwall closure in Section 8.3 depends directly on these lemmas.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Tamed Euler Schemes for Singular SDEs with Multiplicative Levy Noise." pith.science (2026). https://pith.science/paper/JXA3S4JV

@misc{pith2026260718862,
  author       = {Pith},
  title        = {Pith review of: Tamed Euler Schemes for Singular SDEs with Multiplicative Levy Noise},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JXA3S4JV}},
  note         = {Machine review of arXiv:2607.18862}
}
read the original abstract

We prove strong convergence rates for tamed Euler schemes of multidimensional stochastic differential equations with singular drift, multiplicative Brownian noise, and multiplicative Levy noise. The drift is assumed to satisfy a Ladyzhenskaya-Prodi-Serrin type condition, while the diffusion and jump coefficients have Sobolev-type spatial regularity. The proof is based on a nonlocal Zvonkin transform, Krylov-type estimates, and a stochastic Gronwall argument. The main novelty is an error decomposition that isolates two jump-induced contributions: a compensated jump martingale error and a nonlocal compensator error caused by Euler freezing. This decomposition yields explicit rates and recovers the Brownian-type estimate when the jump coefficient vanishes.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

47 extracted references · 6 linked inside Pith

  1. [26]

    L ˆe and C

    K. L ˆe and C. Ling, Taming singular stochastic differential equations: A numerical method, arXiv:2110.01343v6, 2025

    arXiv 2025

  2. [1]

    J. Bao, X. Huang, and S.-Q. Zhang, Convergence rate of EM algorithm for SDEs under integrability condi- tion, arXiv:2009.04781, 2020

    arXiv 2009

  3. [2]

    J. Bao, B. B ¨ottcher, X. Mao, and C. Yuan, Convergence rate of numerical solutions to SFDEs with jumps,J. Comput. Appl. Math., 236(2) (2011), 119–131

    2011

  4. [3]

    Bellingeri, P

    C. Bellingeri, P. K. Friz, and M. Gerencs´er, Singular path spaces and applications,Stoch. Anal. Appl., (2021), 1–24

    2021

  5. [4]

    Butkovsky, K

    O. Butkovsky, K. Dareiotis, and M. Gerencs ´er, Approximation of SDEs: a stochastic sewing approach, Probab. Theory Related Fields, 181 (2021), 975–1034

    2021

  6. [5]

    Chen, C.-S

    P. Chen, C.-S. Deng, R. L. Schilling, and L. Xu, Approximation of the invariant measure of stable SDEs by an Euler–Maruyama scheme,Stochastic Process. Appl., 163 (2023), 136–167

    2023

  7. [6]

    Dareiotis, M

    K. Dareiotis, M. Gerencs ´er, and K. L ˆe, Quantifying a convergence theorem of Gy ¨ongy and Krylov, arXiv:2101.12185, 2021

    arXiv 2021

  8. [7]

    Dareiotis, M

    K. Dareiotis, M. Gerencs ´er, and K. L ˆe, Quantifying a convergence theorem of Gy ¨ongy and Krylov,Ann. Appl. Probab., 33(3) (2023), 2291–2323. 81

    2023

Show all 47 references
  1. [8]

    Dareiotis and M

    K. Dareiotis and M. Gerencs ´er, On the regularisation of the noise for the Euler-Maruyama scheme with irregular drift,Electron. J. Probab., 25 (2020), Paper No. 82

    2020

  2. [9]

    De Angelis, M

    T. De Angelis, M. Germain, and E. Issoglio, A numerical scheme for stochastic differential equations with distributional drift,Stochastic Process. Appl., 154 (2022), 55–90

    2022

  3. [10]

    S. Deng, W. Fei, W. Liu, X. Mao, and S.-Q. Zhang, The truncated EM method for stochastic differential equations with Poisson jumps,J. Comput. Appl. Math., 355 (2019), 232–257

    2019

  4. [11]

    Frazier, B

    M. Frazier, B. Jawerth, and G. Weiss,Littlewood-Paley Theory and the Study of Function Spaces, V ol. 79 of CBMS Regional Conference Series in Mathematics, American Mathematical Society, Providence, RI, 1991

    1991

  5. [12]

    P. K. Friz, A. Hocquet, and K. L ˆe, Rough stochastic differential equations, arXiv:2106.10340, 2021

    2021

  6. [13]

    P. K. Friz and N. B. Victoir,Multidimensional Stochastic Processes as Rough Paths: Theory and Applications, V ol. 120 of Cambridge Studies in Advanced Mathematics, Cambridge University Press, Cambridge, 2010

    2010

  7. [14]

    M. B. Giles, Multilevel Monte Carlo path simulation,Oper. Res., 56(3) (2008), 607–617

    2008

  8. [15]

    Grafakos,Classical Fourier Analysis, V ol

    L. Grafakos,Classical Fourier Analysis, V ol. 2, Springer, New York, 2008

    2008

  9. [16]

    Gy ¨ongy and N

    I. Gy ¨ongy and N. Krylov, Existence of strong solutions for It ˆo’s stochastic equations via approximations, revisited, arXiv:2107.14384, 2021

    2021 arXiv

  10. [17]

    Gy ¨ongy and N

    I. Gy ¨ongy and N. Krylov, Existence of strong solutions for It ˆo’s stochastic equations via approximations, Probab. Theory Related Fields, 105(2) (1996), 143–158

    1996

  11. [18]

    Hutzenthaler, A

    M. Hutzenthaler, A. Jentzen, and P. E. Kloeden, Strong convergence of an explicit numerical method for SDEs with nonglobally Lipschitz continuous coefficients,Ann. Appl. Probab., 22(4) (2012), 1611–1641

    2012

  12. [19]

    Jourdain and S

    B. Jourdain and S. Menozzi, Convergence rate of the Euler-Maruyama scheme applied to diffusion processes withL q−L ρ drift coefficient and additive noise, arXiv:2105.04860, 2021

    2021 arXiv

  13. [20]

    Kaneko and S

    H. Kaneko and S. Nakao, A note on approximation for stochastic differential equations, in:S´ eminaire de Probabilit´ es XXII, Lecture Notes in Math., 1321, Springer, Berlin, 1988, pp. 155–162

    1988

  14. [21]

    Kim,L q(Lp)-theory of parabolic PDEs with variable coefficients,Bull

    K.-H. Kim,L q(Lp)-theory of parabolic PDEs with variable coefficients,Bull. Korean Math. Soc., 45(1) (2008), 169–190

    2008

  15. [22]

    Komatsu, On the martingale problem for generators of stable processes with perturbations,Osaka J

    T. Komatsu, On the martingale problem for generators of stable processes with perturbations,Osaka J. Math., 21 (1984), 113–132

    1984

  16. [23]

    J. H. Kinney, Continuity properties of sample functions of Markov processes,Trans. Amer. Math. Soc., 74 (1953), 280–302

    1953

  17. [24]

    N. V . Krylov,Lectures on Elliptic and Parabolic Equations in Sobolev Spaces, V ol. 96 of Graduate Studies in Mathematics, American Mathematical Society, Providence, RI, 2008

    2008

  18. [25]

    N. V . Krylov and M. R ¨ockner, Strong solutions of stochastic equations with singular time dependent drift, Probab. Theory Related Fields, 131(2) (2005), 154–196

    2005

  19. [27]

    Leoni,A First Course in Sobolev Spaces, American Mathematical Society, Providence, RI, 2017

    G. Leoni,A First Course in Sobolev Spaces, American Mathematical Society, Providence, RI, 2017

    2017

  20. [28]

    Mao, The truncated Euler-Maruyama method for stochastic differential equations,J

    X. Mao, The truncated Euler-Maruyama method for stochastic differential equations,J. Comput. Appl. Math., 290 (2015), 370–384

    2015

  21. [29]

    Meyer,Wavelets and Operators, V ol

    Y . Meyer,Wavelets and Operators, V ol. 37 of Cambridge Studies in Advanced Mathematics, Cambridge University Press, Cambridge, 1992. Translated from the 1990 French original by D. H. Salinger

    1992

  22. [30]

    Mikuleviˇcius and H

    R. Mikuleviˇcius and H. Pragarauskas, On the Cauchy problem for certain integrodifferential operators in Sobolev and H¨older spaces,Lithuanian Math. J., 32(2) (1992), 377–396

    1992

  23. [31]

    Mikuleviˇcius and C

    R. Mikuleviˇcius and C. Zhang, On the rate of convergence of weak Euler approximation for nondegenerate SDEs driven by L´evy processes,Stochastic Process. Appl., 121 (2011), 1720–1748

    2011

  24. [32]

    Mikuleviˇcius and F

    R. Mikuleviˇcius and F. Xu, On the rate of convergence of strong Euler approximation for SDEs driven by L´evy processes,Stochastics, 90(4) (2018), 569–604

    2018

  25. [33]

    Neuenkirch and M

    A. Neuenkirch and M. Sz ¨olgyenyi, The Euler–Maruyama scheme for SDEs with irregular drift: convergence rates via reduction to a quadrature problem,IMA J. Numer. Anal., 41(2) (2021), 1164–1196

    2021

  26. [34]

    O. M. Pamen and D. Taguchi, Strong rate of convergence for the Euler–Maruyama approximation of SDEs with H¨older continuous drift coefficient,Stochastic Process. Appl., 127(8) (2017), 2542–2559

    2017

  27. [35]

    Protter and D

    P. Protter and D. Talay, The Euler scheme for L ´evy driven stochastic differential equations,Ann. Probab., 25(1) (1997), 393–423

    1997

  28. [36]

    T. T. Shieh and D. E. Spector, On a new class of fractional partial differential equations,Adv. Calc. Var., 8 (2014)

    2014

  29. [37]

    Siorpaes, Applications of pathwise Burkholder-Davis-Gundy inequalities,Bernoulli, 24(4B) (2018), 3222–3245

    P. Siorpaes, Applications of pathwise Burkholder-Davis-Gundy inequalities,Bernoulli, 24(4B) (2018), 3222–3245. 82

    2018

  30. [38]

    E. M. Stein,Singular Integrals and Differentiability Properties of Functions, Princeton University Press, Princeton, 1970

    1970

  31. [39]

    Taguchi, A

    D. Taguchi, A. Tanaka, and T. Yuasa,Lq-error estimates for approximation of irregular functionals of random vectors,IMA J. Numer. Anal., (2021), draa096

    2021

  32. [40]

    Triebel,Local Function Spaces, Heat and Navier-Stokes Equations, V ol

    H. Triebel,Local Function Spaces, Heat and Navier-Stokes Equations, V ol. 20 of EMS Tracts in Mathemat- ics, European Mathematical Society (EMS), Z¨urich, 2013

    2013

  33. [41]

    A. Ju. Veretennikov, Strong solutions of stochastic differential equations,Teor. Veroyatnost. i Primenen., 24(2) (1979), 348–360

    1979

  34. [42]

    P. Xia, L. Xie, X. Zhang, and G. Zhao,L q(Lp)-theory of stochastic differential equations,Stochastic Process. Appl., 130(8) (2020), 5188–5211

    2020

  35. [43]

    Xie and X

    L. Xie and X. Zhang, Ergodicity of stochastic differential equations with jumps and singular coefficients, Ann. Inst. Henri Poincar´ e Probab. Stat., 56(1) (2020), 175–229

    2020

  36. [44]

    Zhang, Stochastic homeomorphism flows of SDEs with singular drifts and Sobolev diffusion coefficients, Electron

    X. Zhang, Stochastic homeomorphism flows of SDEs with singular drifts and Sobolev diffusion coefficients, Electron. J. Probab., 16 (2011), 1096–1116

    2011

  37. [45]

    Zhang and G

    X. Zhang and G. Zhao, Heat kernel and ergodicity of SDEs with distributional drifts, arXiv:1710.10537, 2017

    2017 arXiv

  38. [46]

    Zhang, Strong solutions of SDEs with singular drift and Sobolev diffusion coefficients,Stochastic Process

    X. Zhang, Strong solutions of SDEs with singular drift and Sobolev diffusion coefficients,Stochastic Process. Appl., 115 (2005), 1805–1818

    2005

  39. [47]

    A. K. Zvonkin, A transformation of the phase space of a diffusion process that will remove the drift,Mat. Sb. (N.S.), 93(135) (1974), 129–149, 152. 83

    1974

Pith tools

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