pith. sign in

arxiv: 2604.03659 · v3 · submitted 2026-04-04 · 🧮 math.NA · cs.NA

Optimal numerical integration for functions in fractional Gaussian Sobolev spaces

Van Kien Nguyen This is my paper

Pith reviewed 2026-05-13 17:29 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords numerical integrationfractional Sobolev spacesGaussian measuresquadrature rulesHermite expansionsoptimal convergencedominating mixed smoothness
0
0 comments X

The pith

Quadrature rules on the real line achieve the same optimal integration error rates as on the unit cube for fractional Gaussian Sobolev functions.

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

The paper constructs quadrature schemes for integrating functions in fractional Gaussian Sobolev spaces on the full real space by extending known rules from the unit cube. It establishes that these schemes attain the optimal asymptotic order of the integration error for parameters 1 less than p less than infinity and smoothness s greater than one over p but not a natural number. The work further identifies that the spaces at p equals two coincide with Hermite spaces based on weighted summability of Fourier-Hermite coefficients and derives the matching optimal rates for all s above one half. The same optimality result is shown when the spaces are defined through the Gagliardo seminorm.

Core claim

By extending quadrature rules from the fractional Sobolev setting on the unit cube to the Gaussian-weighted space on R^d, the integration error for functions in W^s_p(R^d, γ) attains the optimal asymptotic order in the regime 1 less than p less than infinity and s greater than 1 over p with s not in the natural numbers. The fractional Gaussian Sobolev spaces at p equals two are identical to the Hermite spaces H^s(R^d, γ), which yields the corresponding optimal orders for all s greater than 1/2.

What carries the argument

The quadrature schemes constructed by extending the rules for fractional Sobolev spaces on the unit cube to R^d equipped with the Gaussian measure γ, which carry over the convergence rate exactly.

Load-bearing premise

The extension from quadrature rules on the unit cube to R^d preserves the exact same convergence rate without additional loss from the unbounded domain or the Gaussian weight.

What would settle it

A concrete function in one of the spaces W^s_p(R^d, γ) for which every n-point quadrature rule has an integration error that fails to decay at the claimed optimal asymptotic order.

Figures

Figures reproduced from arXiv: 2604.03659 by Van Kien Nguyen.

Figure 1
Figure 1. Figure 1: Worst-case error of numerical integration for functions in Hermite spaces [PITH_FULL_IMAGE:figures/full_fig_p018_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Worst-case error of numerical integration for functions in Hermite spaces with smooth [PITH_FULL_IMAGE:figures/full_fig_p018_2.png] view at source ↗
read the original abstract

This paper investigates the numerical approximation of integrals for functions in fractional Gaussian Sobolev spaces $W^s_{p}(\mathbb{R}^d,\gamma)$ with dominating mixed smoothness defined via kernel related to the fractional Ornstein-Uhlenbeck operator. Building upon quadrature rules for fractional Sobolev spaces on the unit cube $[-\tfrac{1}{2}, \tfrac{1}{2}]^d$, we construct quadrature schemes on $\mathbb{R}^d$ that achieve the same rate of convergence. As a consequence, we establish the optimal asymptotic order of the integration error in the regime $1 < p < \infty$ and $s > \frac{1}{p}$, $s\not \in \mathbb{N}$. Furthermore, we show that the fractional Gaussian Sobolev spaces $W^s_{2}(\mathbb{R}^d,\gamma)$ coincide with Hermite spaces $\mathcal{H}^s(\mathbb{R}^d,\gamma)$ characterized by the weighted $\ell_2$-summability of their Fourier-Hermite coefficients. From this, we derive the optimal asymptotic order of the integration error for functions in these spaces for all $s > \frac{1}{2}$. We also establish the corresponding optimal asymptotic order for functions in fractional Gaussian Sobolev spaces $W^s_{p,G}(\mathbb{R}^d,\gamma)$ defined via the Gagliardo seminorm.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript constructs quadrature rules for numerical integration of functions belonging to fractional Gaussian Sobolev spaces W^s_p(R^d, γ) on the unbounded domain R^d equipped with the Gaussian measure γ. Building on existing optimal rules for the unit cube, it claims to achieve the optimal convergence rate O(N^{-s/d + 1/(p d)}) for 1 < p < ∞ and s > 1/p with s not integer. It further shows equivalence between W^s_2(R^d, γ) and Hermite spaces H^s(R^d, γ), yielding optimal rates for all s > 1/2, and extends to Gagliardo seminorm definitions.

Significance. If the extension of the quadrature rules preserves the optimal rate without degradation from domain truncation or weight tails, this work would provide important optimal error bounds for integration in Gaussian-weighted fractional Sobolev spaces, relevant for high-dimensional numerical integration in probability and statistics. The equivalence result for p=2 strengthens the connection to classical Hermite expansions and supplies a clean parameter-free derivation in that case.

major comments (2)
  1. [Abstract and §3] Abstract and the extension argument in §3: The claim that cube-based quadrature rules extend to R^d while preserving the exact rate O(N^{-s/d + 1/(p d)}) requires that tail integrals ∫_{|x|>R} f dγ are controlled by the W^s_p(R^d, γ) norm at the same order as the local cube error. The abstract asserts this transfer, but the provided construction does not exhibit explicit tail estimates showing that Gaussian decay introduces neither logarithmic factors nor a reduction in the exponent.
  2. [§4, Theorem 4.2] §4, Theorem 4.2 (Hermite equivalence): The identification of W^s_2(R^d, γ) with H^s(R^d, γ) via weighted ℓ2-summability of Fourier-Hermite coefficients is used to obtain the rate for all s > 1/2. The step from this characterization back to the quadrature error bound on the unbounded domain must confirm that the coefficient decay directly controls the integration error without an extra truncation penalty that would alter the asymptotic order.
minor comments (2)
  1. [Introduction] The distinction between the kernel-based definition of W^s_p(R^d, γ) and the Gagliardo seminorm version W^s_{p,G}(R^d, γ) should be stated explicitly in the introduction, including whether the optimal-rate result holds uniformly for both.
  2. [§2] Notation for the dominating mixed smoothness and the precise definition of the fractional Ornstein-Uhlenbeck kernel could be recalled once more in the statement of the main theorem to improve readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We address each major point below with clarifications on the domain extension and equivalence arguments.

read point-by-point responses

  1. Referee: [Abstract and §3] The claim that cube-based quadrature rules extend to R^d while preserving the exact rate O(N^{-s/d + 1/(p d)}) requires that tail integrals ∫_{|x|>R} f dγ are controlled by the W^s_p(R^d, γ) norm at the same order as the local cube error. The abstract asserts this transfer, but the provided construction does not exhibit explicit tail estimates showing that Gaussian decay introduces neither logarithmic factors nor a reduction in the exponent.

    Authors: In §3 the construction truncates to a cube of radius R(N) chosen to balance truncation and quadrature errors. The Gaussian weight together with the fractional Sobolev norm yields an exponential tail bound on ∫_{|x|>R} |f| dγ that is absorbed into the leading term O(N^{-s/d + 1/(p d)}) without logarithmic factors; the same cutoff argument used for the cube rules carries over directly. We will insert the explicit tail estimates in the revised version. revision: yes

  2. Referee: [§4, Theorem 4.2] The identification of W^s_2(R^d, γ) with H^s(R^d, γ) via weighted ℓ2-summability of Fourier-Hermite coefficients is used to obtain the rate for all s > 1/2. The step from this characterization back to the quadrature error bound on the unbounded domain must confirm that the coefficient decay directly controls the integration error without an extra truncation penalty that would alter the asymptotic order.

    Authors: Theorem 4.2 establishes norm equivalence, so the integration error is controlled by the ℓ2 tail of the Hermite coefficients. Because the quadrature rule is applied after truncation to a finite Hermite expansion whose degree grows with N, the truncation error is bounded by the same coefficient summability and does not change the asymptotic exponent. We will add a short remark after the proof to make this explicit. revision: partial

Circularity Check

0 steps flagged

No significant circularity; extension to R^d is an independent construction

full rationale

The paper builds quadrature schemes for R^d upon previously published rules for the unit cube and claims the same convergence rate O(N^{-s/d + 1/(p d)}) is preserved. It also derives the coincidence of W^s_2(R^d, γ) with Hermite spaces H^s from the kernel definition tied to the fractional Ornstein-Uhlenbeck operator. Neither step reduces a claimed prediction to a fitted parameter, self-definition, or load-bearing self-citation within this manuscript; the extension and equivalence are presented as new results whose validity rests on the explicit construction and norm equivalences rather than tautological renaming or internal fitting. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, ad-hoc axioms, or invented entities are visible. The work rests on standard properties of fractional Sobolev spaces and known quadrature results on the cube.

axioms (1)
  • standard math Quadrature rules for fractional Sobolev spaces on the unit cube achieve the stated rates
    The paper explicitly builds upon these rules to extend to R^d.

pith-pipeline@v0.9.0 · 5540 in / 1134 out tokens · 41452 ms · 2026-05-13T17:29:21.916655+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

32 extracted references · 32 canonical work pages

  1. [1]

    N. S. Bakhvalov. Optimal convergence bounds for quadrature processes and integration methods of monte carlo type for classes of functions.Zhurnal Vychislitel’noi Matematiki i Matematicheskoi Fiziki, 4:5–63, 1963

    work page 1963

  2. [2]

    Carbotti, S

    A. Carbotti, S. Cito, D. A. L. Manna, and D. Pallara. Gamma convergence of Gaussian fractional perimeter.Adv. Calc. Var., 16:571–595, 2023

    work page 2023

  3. [3]

    D˜ ung and V

    D. D˜ ung and V. K. Nguyen. Optimal numerical integration and approximation of functions onR d equipped with Gaussian measure.IMA J. Numer. Anal., 44:1242––1267, 2024

    work page 2024

  4. [4]

    D˜ ung, V

    D. D˜ ung, V. N. Temlyakov, and T. Ullrich.Hyperbolic Cross Approximation. Advanced Courses in Mathematics - CRM Barcelona, Birkh¨ auser/Springer, 2018

    work page 2018

  5. [5]

    J. Dick, C. Irrgeher, G. Leobacher, and F. Pillichshammer. On the optimal order of in- tegration in Hermite spaces with finite smoothness.SIAM J. Numer. Anal., 56:684–707, 2018

    work page 2018

  6. [6]

    Dick and F

    J. Dick and F. Pillichshammer.Digital Nets and Sequences: Discrepancy Theory and Quasi- Monte Carlo Integration. Cambridge University Press, Cambridge, 2010

    work page 2010

  7. [7]

    V. Dubinin. Cubature formulas for classes of functions with bounded mixed difference. Sbornik: Mathematics, 76:283–292, 1993

    work page 1993

  8. [8]

    V. Dubinin. Cubature formulae for Besov classes.Izvestiya: Mathematics, 61:259–283, 1997

    work page 1997

  9. [9]

    Dung and T

    D. Dung and T. Ullrich. Lower bounds for the integration error for multivariate functions with mixed smoothness and optimal fibonacci cubature for functions on the square.Math. Nachr., 288:743–762, 2015

    work page 2015

  10. [10]

    K. K. Frolov. Upper error bounds for quadrature formulas on function classes.Doklady Akademii Nauk SSSR, 231:818–821, 1976

    work page 1976

  11. [11]

    Hansen and J

    M. Hansen and J. Vybiral. The jawerth–franke embedding of spaces of dominating mixed smoothness.Georgian Math. J., 16(4):667–682, 2009

    work page 2009

  12. [12]

    Hinrichs

    A. Hinrichs. Discrepancy of Hammersley points in Besov spaces of dominating mixed smoothness.Math. Nachr., 283:478–488, 2010

    work page 2010

  13. [13]

    Hinrichs, L

    A. Hinrichs, L. Markhasin, J. Oettershagen, and T. Ullrich. Optimal quasi-monte carlo rules on order 2 digital nets for the numerical integration of multivariate periodic functions. Numer. Math., 134:163–196, 2014

    work page 2014

  14. [14]

    E. Hlawka. Zur angen¨ aherten Berechnung mehrfacher Integrale.Monatsh. Math., 66:140– 151, 1962

    work page 1962

  15. [15]

    Irrgeher, P

    C. Irrgeher, P. Kritzer, G. Leobacher, and F. Pillichshammer. Integration in Hermite spaces of analytic functions.J. Complexity, 31:308–404, 2015

    work page 2015

  16. [16]

    Irrgeher and G

    C. Irrgeher and G. Leobacher. High-dimensional integration on theR d, weighted Hermites- paces, and orthogonal transforms.J. Complexity, 31:174–205, 2015. 19

    work page 2015

  17. [17]

    N. M. Korobov. Approximate evaluation of repeated integrals.Dokl. Akad. Nauk SSSR, 124, 1959

    work page 1959

  18. [18]

    Krieg and E

    D. Krieg and E. Novak. A universal algorithm for multivariate integration.Found. Comput. Math., 17:895–916, 2016

    work page 2016

  19. [19]

    Lunardi, G

    A. Lunardi, G. Metafune, and D. Pallara. The Ornstein-Uhlenbeck semigroup in finite dimensions.Philos. Trans. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci., 378, 2020

    work page 2020

  20. [20]

    Markhasin

    L. Markhasin. Discrepancy and integration in function spaces with dominating mixed smoothness.Diss. Math., 494:1–81, 2013

    work page 2013

  21. [21]

    V. K. Nguyen and W. Sickel. Pointwise multipliers for Sobolev and Besov spaces of domi- nating mixed smoothness.J. Math. Anal. Appl., 452:62–90, 2017

    work page 2017

  22. [22]

    V. K. Nguyen, M. Ullrich, and T. Ullrich. Change of variable in spaces of mixed smoothness and numerical integration of multivariate functions on the unit cube.Constr. Approx., 46:69–108, 2017

    work page 2017

  23. [23]

    Novak and H

    E. Novak and H. Wo´ zniakowski.Tractability of Multivariate Problems, Volume II: Standard Information for Functionals. EMS Tracts in Mathematics, Vol. 12, Eur. Math. Soc. Publ. House, Z¨ urich, 2010

    work page 2010

  24. [24]

    Schmeisser and H

    H. Schmeisser and H. Triebel.Topics in Fourier Analysis and Function Spaces. Chichester; New York : Wiley, 1987

    work page 1987

  25. [25]

    Szeg¨ o.Orthogonal Polynomials

    G. Szeg¨ o.Orthogonal Polynomials. Amer. Math. Soc., 1939

    work page 1939

  26. [26]

    V. N. Temlyakov. Approximation of functions with bounded mixed derivative.Trudy MIAN, 178:1–112, 1986

    work page 1986

  27. [27]

    V. N. Temlyakov. On a way of obtaining lower estimates for the errors of quadrature formulas.Matem. Sb., pages 1403–1413, 1990

    work page 1990

  28. [28]

    V. N. Temlyakov. Error estimates for Fibonacci quadrature formulas for classes of functions with a bounded mixed derivative.Trudy Mat. Inst. Steklov, 200:327–335, 1991

    work page 1991

  29. [29]

    V. N. Temlyakov.Approximation of Periodic Functions. Computational Mathematics and Analysis Series, Nova Science Publishers, Inc., Commack, NY., 1993

    work page 1993

  30. [30]

    V. N. Temlyakov. Cubature formulas, discrepancy, and nonlinear approximation.J. Com- plexity, 19, 2003

    work page 2003

  31. [31]

    Triebel.Bases in Function Spaces, Sampling, Discrepancy, Numerical Integration

    H. Triebel.Bases in Function Spaces, Sampling, Discrepancy, Numerical Integration. EMS Tracts in Mathematics, European Mathematical Society (EMS), Z¨ urich, 2010

    work page 2010

  32. [32]

    Ullrich and T

    M. Ullrich and T. Ullrich. The role of Frolov’s cubature formula for functions with bounded mixed derivative.SIAM J. Numer. Anal., 54, No. 2:969–993, 2016. 20

    work page 2016