Pith. sign in

REVIEW 3 major objections 7 minor 34 references

Error Analysis of the Deep Mixed Residual Method for High-order Elliptic Equations

T0 review · 3 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper proves a priori error bounds for the deep mixed residual method showing that two-layer ReQU and ReCU networks solve high-order elliptic equations at dimension-exempt rates, provided the exact solution lies in a spectral Barron…

desk verdict First-order MIM bounds are credible and new; the second-order theorem has a real but repairable gap—an H^1 approximation lemma where H^2 is required. read the letter →

arxiv 2411.14151 v2 pith:RS2JYCP5 submitted 2024-11-21 math.NA cs.NA

Mengjia Bai , Jingrun Chen , Rui Du , Zhiwei Sun This is my paper · ORCID
classification math.NAcs.NA MSC 65N1568Q25
keywords deepmixedresidualmethodhigh-orderellipticequationsapriorierroranalysisBarronspaceRademachercomplexitycoercivitycurseofdimensionalityleastsquaressystems
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 establishes a priori error estimates for the deep mixed residual method (MIM) applied to $2n$-th order elliptic equations on the unit cube, with non-homogeneous Dirichlet, Neumann, or Robin boundary data. The central claim is that when the exact solution lies in a spectral Barron space, the expected squared error of the MIM approximations to the solution and its Laplacian derivatives is bounded by $O(\|u^*\|_{B}^2/m + \|u^*\|_{B}^2/\sqrt N)$ for the Neumann and Robin cases, and by $O(\|u^*\|_{B}^2/\sqrt m + \|u^*\|_{B}^2/N^{1/4})$ for the Dirichlet case; here $m$ is the network width and $N$ the number of interior collocation points. The constants are at most polynomial in the dimension, so the bound does not suffer the curse of dimensionality. The proof splits the error into approximation, generalization, and optimization components, using Céa's lemma, and controls the first two by Barron-space approximation and Rademacher complexity respectively. A side consequence is that MIM needs only quadratic or cubic ReLU activations, whereas the deep Ritz method needs ReLU of order $n+1$ for the same equation.

What carries the argument

The load-bearing object is the bilinear form $B_\alpha(u,w)=(Pu,Pw)_{L^2(\Omega)}+\lambda(S_\alpha u,S_\alpha w)_{L^2(\partial\Omega)}$ (plus a zero-mean integral term for Neumann), where $P$ is a first-order differential operator coupling $\nabla\phi_k$ to $\psi_k$ and $\mathrm{div}\,\psi_k$ to $\phi_{k+1}$, and $S_\alpha$ is the boundary trace operator for Dirichlet, Neumann, or Robin data; the second-order system uses the analogous operator $P^*$ with $\Delta\phi_k-\phi_{k+1}$. Coercivity of $B_\alpha$ in the $H^1\times H(\mathrm{div})$ norm is proven by a perturbation technique: a sequence of small parameters $\delta_k$ weights the cross terms so that Young's inequality absorbs them into the decoupled energy. The Dirichlet case, where coercivity in $H^1$ is impossible, uses a sup-linear coercivity estimate $\sqrt{B_D(u,u)}\,E \gtrsim \|u\|^2_Z$, which after Céa's lemma turns the rates into square roots.

What would settle it

Construct a sequence of exact solutions of the elliptic equation whose Barron norm grows with the dimension but whose classical Sobolev regularity is fixed, and run MIM with ReQU networks: if the measured expected $H^1\times H(\mathrm{div})$ error does not decay at least as $C(1/m+1/\sqrt N)$ with $C$ polynomial in $d$, the claimed dimension-exempt rate fails. Concretely, take $u^*=\cos(\pi x_1)$ on $[0,1]^d$ with Neumann boundary data, train the first-order system with increasing $m$ and $N$, and check that the squared error obeys the predicted $1/m+1/\sqrt N$ scaling; a slower decay for fixed $d$ contradicts the theorem.

Watch

Extended reading notes

Core claim

The paper proves that the MIM empirical risk minimizer converges to the true solution tuple $(\phi_k,\psi_k)\approx(\Delta^k u^*,\nabla\Delta^k u^*)$ at a rate that is additive in the approximation and generalization errors. For the first-order least-squares system with ReQU networks, assuming $u^*\in B^{2n+2}$, the expectation over Monte Carlo samples satisfies $\sum_{k=0}^{n-1}(E\|\hat\phi_k-\Delta^k u^*\|^2_{H^1} + E\|\hat\psi_k-\nabla\Delta^k u^*\|^2_{H(\mathrm{div})}) \lesssim \|u^*\|^2_{B^{2n+2}}/m + \|u^*\|^2_{B^{2n+2}}/\sqrt N$ for Neumann and Robin conditions. For Dirichlet boundary data, the standard $L^2$ penalty causes a loss of $3/2$ in Sobolev regularity, so the paper derives a sup-linear coercivity for the bilinear form and obtains the square-root rate $\sqrt{E_{\mathrm{app}}}+\sqrt{E_{\mathrm{gen}}}$, giving $O(1/\sqrt m + N^{-1/4})$. The second-order system with ReCU networks repeats the same structure under the stronger assumption $u^*\in B^{2n+3}$, yielding analogous estimates for $\hat\phi_k-\Delta^k u^*$ in $H^1$ and $\nabla\hat\phi_k-\nabla\Delta^k u^*$ in $H(\mathrm{div})$.

Load-bearing premise

The exact solution is assumed to belong to a spectral Barron space $B^{2n+2}$ or $B^{2n+3}$, and the proof uses that assumption both for the $1/m$ approximation rate and for the a priori parameter bounds in the network class; if a solution of the elliptic equation is not Barron-regular, the dimension-exempt conclusion has no basis here.

Editorial extensions

If this is right

  • For Neumann and Robin boundary data, the theorem predicts that the expected squared $H^1(\Omega)\times H(\mathrm{div};\Omega)$ error over all Laplacian iterates decays as $O(\|u^*\|^2_B/m + \|u^*\|^2_B/\sqrt N)$, with an additional boundary sample count $N_\partial = O(N^{d/2})$ keeping the boundary contribution from spoiling this rate.
  • For Dirichlet data the same error decays only as $O(\|u^*\|^2_B/\sqrt m + \|u^*\|^2_B/N^{1/4})$, so the boundary penalty costs a square root in both network width and sample count.
  • MIM requires only ReQU (quadratic) activations for the first-order formulation and ReCU (cubic) activations for the second, while the deep Ritz method needs $\mathrm{ReLU}^{n+1}$; for high-order equations this is a large reduction in required activation smoothness.
  • The bounds are statements about expectation with respect to the random training samples and about global minimizers of the empirical loss; optimization error is not analyzed, so a training procedure that fails to reach the global minimum can produce larger errors than the theorem bounds.

Reading between the lines

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

  • An implication left implicit in the paper is that the practical rate of MIM depends on the optimizer's ability to reach a global minimizer of the nonconvex empirical loss, so the theorem is best read as a guarantee on the landscape rather than on gradient descent.
  • Because the rate improvement over deep Ritz comes from the mixed formulation's lower-order loss, the same Barron/Céa/Rademacher decomposition should apply to other mixed least-squares neural methods, for instance first-order systems for elasticity or Stokes flow.
  • The Dirichlet square-root rate suggests that any penalty-free exact boundary enforcement in the spirit of the original MIM papers might restore the $O(1/m)$ rate, but the present analysis deliberately works with the $L^2$ penalty and does not make that claim.
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 / 7 minor

Summary. The paper analyzes the Deep Mixed Residual Method (MIM) for 2n-order elliptic equations with non-homogeneous Dirichlet, Neumann, and Robin boundary conditions. Two least-squares formulations are studied: a first-order system approximated by two-layer ReQU networks and a second-order system approximated by two-layer ReCU networks. Using spectral Barron spaces, a Céa-lemma framework from [32], perturbation-based coercivity estimates, and Rademacher complexity bounds, the authors claim a priori error estimates of the form O(||u*||^2/m + ||u*||^2/N^{1/2}) for Neumann/Robin and O(||u*||^2/m^{1/2} + ||u*||^2/N^{1/4}) for Dirichlet, with constants depending at most polynomially on dimension. The proof consists of coercivity lemmas, approximation lemmas for Barron functions, generalization bounds, and an error decomposition; the optimization error is explicitly left out of scope.

Significance. If the results were fully supported, the paper would make a valuable contribution: it extends the MIM error analysis to high-order equations with non-homogeneous boundary conditions, gives dimension-exempt rates in terms of network width m and sample sizes N, and identifies that MIM lowers the activation-function regularity relative to the deep Ritz method. The perturbation argument with the carefully chosen parameter sequence delta_k and the resulting coercivity estimates for the coupled first-order and second-order systems are technically substantial. The treatment of non-homogeneous Neumann and Robin conditions, including zero-mean penalties, is also a strength. However, the second-order theorem as written is not derivable from the stated lemmas, and the Dirichlet sup-linear coercivity step needs clarification; these are load-bearing gaps that prevent the paper from being accepted in its current form.

major comments (3)
  1. [Section 3.3, Eq. (38), Lemma 3.7, and Appendix A.1] Theorem 2.2 is not supported by the stated approximation estimate. The Céa step in Lemma 3.5 requires E*_app = inf over FReCU,m of sum_k ||phi_k - Delta^k u*||^2_{H^2(Omega)}, because the second-order bilinear form is bounded on (H^2)^n and the Neumann/Robin traces involve normal derivatives. However, Lemma 3.7 only establishes an H^1 approximation rate, and its proof in Appendix A.1 (Lemma A.2 and the construction around Eq. (92)) yields ||g - g_m||^2_{H^1} <= C B^2/m with no control of Delta(g - g_m) in L^2. Since H^1 closeness does not imply H^2 closeness, the bound on E*_app does not follow, and the H(Omega)-type error estimate in Theorem 2.2 cannot be concluded. A repair requires either a new H^2 approximation lemma for ReCU networks or a reformulation of the second-order loss/coercivity in weaker norms.
  2. [Section 3.1.1, Lemma 3.1, and Section 4.4, Eq. (48)] The Dirichlet sup-linear coercivity statement needs correction or clarification. The proof of Lemma 4.1 establishes the homogeneous inequality (B_D^{1/2}(u,u) + sum_k ||psi_k||_{L^2(dOmega)}) B_D^{1/2}(u,u) >= C sum_k (||phi_k||^2_{H^1} + ||psi_k||^2_{H(div)}). When ||u||_{H^1} <= E is imposed, the trace and boundedness estimates give B_D^{1/2}(u,u) >= (c/E) sum_k (||phi_k||^2_{H^1} + ||psi_k||^2_{H(div)}), not c E times that sum as displayed. The factor E^{-1} is essential: it must be propagated through Céa's lemma to obtain the scaling ||u*||^2/m^{1/2} in the Dirichlet parts of Theorems 2.1 and 2.2. Please correct the statement of Lemma 3.1 and indicate explicitly how E enters the constants in Lemma 3.3 and the final estimates.
  3. [Section 3.4, Lemma 3.10, and Appendix A.2] The generalization bound for the second-order Neumann and Robin cases is incomplete. For alpha = N, R, the boundary loss classes S*_alpha contain normal derivatives partial_n phi_k. The proof of Lemma 3.10 estimates R_hatN(S*_alpha) using only R_hatN(FReCU,m) and a supremum bound on ||S*_alpha v_theta - g_alpha||_Linfty, but partial_n v_theta is not an element of FReCU,m, so Lemma A.5 does not apply to it. A separate Rademacher bound for the derivative network class {partial_n v_theta : v_theta in FReCU,m} is needed to justify the N^{-1/2} boundary contribution in Theorem 2.2 for Neumann and Robin conditions.
minor comments (7)
  1. [Theorem 2.2] The statement contains a typo: the error bound is written with ||phi*_alpha||^2_{B^{2n+3}} but should refer to ||u*_alpha||^2_{B^{2n+3}}.
  2. [Section 3.2.2] The text says 'by applying Lemma 3.1 and Céa Lemma' in the second-order case; the correct reference is Lemma 3.2.
  3. [Eq. (2) versus Eq. (14)] The original loss in Eq. (2) displays the terms ||phi_{i+1} - div psi_i||^2 with an L^2(dOmega) boundary norm, whereas the analyzed first-order loss in Eq. (14) uses the interior L^2(Omega) norm through the operator P. Please clarify the discrepancy.
  4. [Theorem 2.1 and Theorem 2.2] The Dirichlet rate is printed as ||u*||^2/(4\sqrt{N}), which is easy to misread as a fourth root only from the text; use \sqrt[4]{N} consistently.
  5. [After Eq. (9)] The paper assumes u* belongs to B^{2n+2} or B^{2n+3} and asserts that elliptic regularity theory can guarantee this for smooth data, but no theorem or argument is supplied. Since the m^{-1} approximation rate and the dimension-exemption conclusion fail if u* is not in the relevant Barron space, this assumption should be stated explicitly as a regularity condition rather than as a consequence of standard elliptic regularity.
  6. [Eqs. (10) and (18)] The network classes are defined using the unknown norm ||u*|| as an a priori bound on the parameters, making the approximation and generalization statements target-aware. This is acceptable as an a priori estimate, but it should be stated that the admissible set is not known in practice.
  7. [Section 2.3.1, Neumann loss] For the Neumann problem, the exact solution is unique only up to a constant; the paper should specify that u*_N denotes the unique zero-mean solution, so that the zero-mean penalty in the loss does not change the minimizer.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the MIM error bound is obtained from external Barron/Rademacher lemmas plus proved coercivity estimates; the target-dependence of the network class is an a priori caveat, not an input-output equivalence.

full rationale

The paper's derivation chain is standard and non-circular: coercivity and boundedness of the bilinear forms (Section 3.1, proved in Sections 4 and 5), Céa's Lemma decomposition (Lemma 3.3), Barron-space approximation (Lemmas 3.6 and 3.7), and Rademacher-complexity generalization bounds (Lemmas 3.8-3.10). The approximation rates ||u*||^2/m and the generalization rates ||u*||^2/sqrt(N) are quoted from or derived via external results, chiefly [13] and [16], not from the paper's own target theorems. The network classes in (10) and (18) are defined with a priori parameter bounds involving the unknown solution norm ||u*||_{B^s}; this makes the estimates non-implementable and target-aware, but it is not circular because the class definition does not already contain the error bound that the approximation lemma proves. The self-citations to the authors' earlier MIM papers [17,18] are background references to the method being analyzed and are not load-bearing in the error estimates. I note a genuine, non-circular correctness gap in Theorem 2.2: Lemma 3.7 provides only an H^1 approximation for ReCU networks, whereas the proof of Lemma 3.5 and Eq. (38) require E*_app in the H^2 norm; the second-order theorem is therefore not supported by the stated lemmas. This is a missing estimate or proof gap, not an equivalence between the paper's inputs and outputs, so it does not raise the circularity score.

Assumptions & free parameters 3 free parameters · 7 assumptions · 0 invented entities

The new content is the perturbation coercivity and its assembly into Cea estimates. Everything else is either standard functional analysis or external Barron/Rademacher theory, plus a strong Barron regularity assumption on the solution and the sampling-budget assumption N_∂ = O(N^{d/2}).

free parameters (3)
  • lambda (boundary penalty weight)
    Boundary penalty weight in losses (14), (16), (22), (24); coercivity requires lambda sufficiently large, e.g., lambda - C*delta > 0 in the proofs of Lemmas 4.1 and 5.1.
  • mu (zero-mean penalty weight)
    Zero-mean penalty for Neumann boundary in (16), (24); requires mu sufficiently large to absorb the Poincare constant terms.
  • delta (perturbation parameter) = small positive, not fitted
    Introduced in Eq. (53) to control cross terms in the coercivity proofs. It is a proof device, not part of the final algorithm, and the final constants absorb it.
assumptions (7)
  • domain assumption Solution u* lies in spectral Barron space B^{2n+2} or B^{2n+3}; see after Eq. (9).
    All approximation rates rest on Barron regularity. The paper justifies this by a reference to elliptic regularity rather than by a proof.
  • domain assumption Well-posedness of (4)-(7) on Omega = [0,1]^d with non-homogeneous Dirichlet, Neumann, and Robin data.
    The analysis assumes a unique solution with sufficient regularity to identify the mixed-system target, used in Sections 2.1 and 3.2.
  • standard math Standard trace inequalities (49)-(50), Poincare-Friedrichs inequalities (51)-(52), and elliptic regularity estimates (81), (83).
    Used throughout the coercivity proofs in Sections 4 and 5.
  • standard math Barron approximation theorems for ReQU and ReCU two-layer networks (Lemmas 3.6, 3.7, A.1, A.2) cited from [13,14,16].
    These external results supply the 1/m approximation-error bound in Barron norm.
  • standard math Rademacher complexity bounds for linear and ReCU classes (Lemmas A.4, A.5) and loss contraction properties from [13].
    Used for the generalization-error estimate 1/sqrt(N).
  • standard math Cea's Lemma framework for bilinear forms from [32] (Lemma 3.3).
    Provides the abstract error decomposition into approximation, generalization, and optimization errors.
  • domain assumption Boundary sample budget satisfies N_∂ = O(N^{d/2}).
    Used in Lemmas 3.9 and 3.10 to keep the boundary Rademacher complexity at 1/sqrt(N); in high dimension this is a severe sampling cost.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Error Analysis of the Deep Mixed Residual Method for High-order Elliptic Equations." pith.science (2026). https://pith.science/paper/RS2JYCP5

@misc{pith2026241114151,
  author       = {Pith},
  title        = {Pith review of: Error Analysis of the Deep Mixed Residual Method for High-order Elliptic Equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RS2JYCP5}},
  note         = {Machine review of arXiv:2411.14151}
}
read the original abstract

This paper presents an a priori error analysis of the Deep Mixed Residual method (MIM) for solving high-order elliptic equations with non-homogeneous boundary conditions, including Dirichlet, Neumann, and Robin conditions. We examine MIM with two types of loss functions, referred to as first-order and second-order least squares systems. By providing boundedness and coercivity analysis, we leverage C\'{e}a's Lemma to decompose the total error into the approximation, generalization, and optimization errors. Utilizing the Barron space theory and Rademacher complexity, an a priori error is derived regarding the training samples and network size that are exempt from the curse of dimensionality. Our results reveal that MIM significantly reduces the regularity requirements for activation functions compared to the deep Ritz method, implying the effectiveness of MIM in solving high-order equations.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 33 canonical work pages

  1. [32]

    A Unified Framework for the Error Analysis of Physics-Informed Neural Networks

    Marius Zeinhofer, Rami Masri, and Kent-Andr´ e Mardal, A unified framework for the error analysis of physics-informed neural networks , arXiv preprint arXiv:2311.00529 (2023). Appendix A. Neural-Network Approximation A.1. The proof of Lemma 3.7. For simplicity, we will only provide the proof for the case when the output function is one-dimension al here. T...

    work page Pith review arXiv 2023

  2. [1]

    2, 153–180

    Lars-Erik Andersson, Tommy Elfving, and Gene H Golub, Solution of biharmonic equations with application to radar imaging , Journal of Computational and Applied Mathematics 94 (1998), no. 2, 153–180

    work page 1998

  3. [2]

    3, 930–945

    Andrew R Barron, Universal approximation bounds for superpositions of a sig moidal function, IEEE Transactions on Information theory 39 (1993), no. 3, 930–945

    work page 1993

  4. [3]

    Zhiqiang Cai, Jingshuang Chen, Min Liu, and Xinyu Liu, Deep least-squares methods: An unsupervised learning-based numerical method for solvi ng elliptic PDEs , Journal of Computational Physics 420 (2020), 109707

    work page 2020

  5. [4]

    6, 1785–1799

    Zhiqiang Cai, R Lazarov, Thomas A Manteuffel, and Stephen F McCormick, First- order system least squares for second-order partial differen tial equations: Part I , SIAM Journal on Numerical Analysis 31 (1994), no. 6, 1785–1799

    work page 1994

  6. [5]

    38, 385704

    WH Duan and Chien Ming Wang, Exact solutions for axisymmetric bending of micro/nanoscale circular plates based on nonlocal plate th eory, Nanotechnology 18 (2007), no. 38, 385704

    work page 2007

  7. [6]

    Weinan E and Bing Yu, The deep ritz method: a deep learning-based numerical algo- rithm for solving variational problems , Communications in Mathematics and Statistics 6 (2018), no. 1, 1–12. 32 MENGJIA BAI, JINGRUN CHEN, RUI DU, AND ZHIWEI SUN

    work page 2018

  8. [7]

    159, Springer, 2004

    Alexandre Ern and Jean-Luc Guermond, Theory and practice of finite elements , vol. 159, Springer, 2004

    work page 2004

Show all 34 references
  1. [8]

    5, Springer Science & Business Media, 2012

    Vivette Girault and Pierre-Arnaud Raviart, Finite element methods for Navier–Stokes equations: theory and algorithms , vol. 5, Springer Science & Business Media, 2012

    2012

  2. [9]

    34, 8505–8510

    Jiequn Han, Arnulf Jentzen, and Weinan E, Solving high-dimensional partial differen- tial equations using deep learning , Proceedings of the National Academy of Sciences 115 (2018), no. 34, 8505–8510

    2018

  3. [10]

    Qingguo Hong, Jonathan W Siegel, and Jinchao Xu, A priori analysis of stable neural network solutions to numerical PDEs , arXiv preprint arXiv:2104.02903 (2021)

    2021 arXiv

  4. [11]

    5, 051129

    Evgeniy Khain and Leonard M Sander, Generalized Cahn-Hilliard equation for biolog- ical applications, Physical Review E—Statistical, Nonlinear, and Soft Matte r Physics 77 (2008), no. 5, 051129

    2008

  5. [12]

    7, Elsevier, 2012

    Lev Davidovich Landau, LP Pitaevskii, Arnol’d Markovic h Kosevich, and Ev- genii Mikhailovich Lifshitz, Theory of elasticity: volume 7 , vol. 7, Elsevier, 2012

    2012

  6. [13]

    Lingfeng Li, Xue-Cheng Tai, Jiang Yang, and Quanhui Zhu, A priori error estimate of deep mixed residual method for elliptic PDEs , Journal of Scientific Computing 98 (2024), no. 2, 44

    2024

  7. [14]

    1, 319–344

    Yuanyuan Li, Shuai Lu, Peter Math´ e, and Sergei V Perever zev, Two-layer networks with the ReLU k activation function: Barron spaces and derivative approxi mation, Numerische Mathematik 156 (2024), no. 1, 319–344

    2024

  8. [15]

    Jacques Louis Lions and Enrico Magenes, Non-homogeneous boundary value problems and applications: Vol. 1 , vol. 181, Springer Science & Business Media, 2012

    2012

  9. [16]

    Yulong Lu, Jianfeng Lu, and Min Wang, A priori generalization analysis of the deep ritz method for solving high dimensional elliptic partial d ifferential equations , Confer- ence on learning theory (2021), 3196–3241

    2021

  10. [17]

    4, 748–775

    Liyao Lyu, Keke Wu, Rui Du, and Jingrun Chen, Enforcing exact boundary and initial conditions in the deep mixed residual method , CSIAM Transactions on Applied Mathematics 2 (2021), no. 4, 748–775

    2021

  11. [18]

    Liyao Lyu, Zhen Zhang, Minxin Chen, and Jingrun Chen, MIM: A deep mixed residual method for solving high-order partial differential equation s, Journal of Computational Physics 452 (2022), 110930

    2022

  12. [19]

    2, 981–1022

    Siddhartha Mishra and Roberto Molinaro, Estimates on the generalization error of physics informed neural networks (PINNs) for approximatin g a class of inverse prob- lems for PDEs , IMA Journal of Numerical Analysis 42 (2022), no. 2, 981–1022

    2022

  13. [20]

    , Estimates on the generalization error of physics-informed neural networks for approximating PDEs , IMA Journal of Numerical Analysis 43 (2023), no. 1, 1–43

    2023

  14. [21]

    Johannes M¨ uller and Marius Zeinhofer, Error estimates for the variational training of neural networks with boundary penalty , arXiv preprint arXiv:2103.01007 (2021)

    2021 arXiv

  15. [22]

    , Error estimates for the deep ritz method with boundary penal ty, Mathematical and Scientific Machine Learning (2022), 215–230

    2022

  16. [23]

    , Notes on exact boundary values in residual minimisation , Mathematical and Scientific Machine Learning (2022), 231–240

    2022

  17. [24]

    Jindˇ rich Neˇ cas,Les m´ ethodes directes en th´ eorie des ´ equations elliptiques, Academia (1967)

    1967

  18. [25]

    Maziar Raissi, Paris Perdikaris, and George E Karniadak is, Physics-informed neural networks: A deep learning framework for solving forward and inverse problems in- volving nonlinear partial differential equations , Journal of Computational physics 378 (2019), 686–707

    2019

  19. [26]

    1, 47–69

    Martin Schechter, On Lp estimates and regularity II , Mathematica Scandinavica 13 (1963), no. 1, 47–69

    1963

  20. [27]

    ERROR ANALYSIS OF MIM FOR HIGH-ORDER ELLIPTIC EQUATIONS 33

    Jonathan W Siegel, Qingguo Hong, Xianlin Jin, Wenrui Hao , and Jinchao Xu, Greedy training algorithms for neural networks and applications t o PDEs , Journal of Compu- tational Physics 484 (2023), 112084. ERROR ANALYSIS OF MIM FOR HIGH-ORDER ELLIPTIC EQUATIONS 33

    2023

  21. [28]

    Justin Sirignano and Konstantinos Spiliopoulos, DGM: A deep learning algorithm for solving partial differential equations , Journal of computational physics 375 (2018), 1339–1364

    2018

  22. [29]

    2, 04021154

    Mohammad Vahab, Ehsan Haghighat, Maryam Khaleghi, and N asser Khalili, A physics-informed neural network approach to solution and i dentification of biharmonic equations of elasticity , Journal of Engineering Mechanics 148 (2022), no. 2, 04021154

    2022

  23. [30]

    55 (2022), no

    E Weinan, Chao Ma, and Lei Wu, The Barron space and the flow-induced function spaces for neural network models , Constructive Approximation. 55 (2022), no. 1, 369– 406

    2022

  24. [31]

    Yaohua Zang, Gang Bao, Xiaojing Ye, and Haomin Zhou, Weak adversarial networks for high-dimensional partial differential equations , Journal of Computational Physics 411 (2020), 109409

    2020

  25. [33]

    Let us define the function class: Fcos(B) := { γ 1 + π4|k|4 1 cos ( π(k ·x) + b ) ⏐ ⏐ ⏐k ∈ Zd/{0}, |γ| ≤ B, b ∈ {0, 1} } , where B > 0 is a constant

    ≤ ‖u‖B4(Ω) , and g(x, k) = Zu 1 + π4|k|4 1 · 1 2d ∑ ξ∈Ξ cos ( π(kξ ·x + θk) ) , with θ(k) ∈ {0, 1} andkξ = (k1ξ1, · · · kdξd). Let us define the function class: Fcos(B) := { γ 1 + π4|k|4 1 cos ( π(k ·x) + b ) ⏐ ⏐ ⏐k ∈ Zd/{0}, |γ| ≤ B, b ∈ {0, 1} } , where B > 0 is a constant. H...

  26. [34]

    It follows that from property ( 5) (98) RN (L∗) ≤4 sup vθ ∈FReCU,m ‖P∗vθ −f ‖L∞ (Ω) ( RN (f ) + m∑ i=1 d∑ j=1 |ai||Wi,j|2RN ( σ(2)(G) ) )

    of P ∗, it follows that (97) |P∗vθ −f |2 = |∆ ϕn−1 − f |2 + n−2∑ i=0 |ϕi+1 − ∆ ϕi|2. It follows that from property ( 5) (98) RN (L∗) ≤4 sup vθ ∈FReCU,m ‖P∗vθ −f ‖L∞ (Ω) ( RN (f ) + m∑ i=1 d∑ j=1 |ai||Wi,j|2RN ( σ(2)(G) ) ) . According to the definition of the neural network and...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.