Pith. sign in

REVIEW 3 major objections 4 minor 42 references

Multiscale Mixed-Dimensional Simulation via Domain Decomposition and Non-Intrusive Neural Model Order Reduction

T0 review · 3 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 The paper claims that replacing the exact local operators of a two-level Robin-Robin domain-decomposition iteration with neural surrogates preserves the contraction mechanism, and proves the asymptotic global error stays bounded by CDD/(1-q

desk verdict A solid, subfield-level contribution that replaces all fine-scale local operations in a two-level Robin-Robin DD iteration with cascade-trained neural surrogates; the theory is standard and the main caveat is that the uniform-accuracy premise is asserted rather than verified in the norm the bound actually needs. read the letter →

arxiv 2607.15171 v2 pith:B43WM74N submitted 2026-07-16 math.NA cs.NA

Nunzio Dimola , Piermario Vitullo , Paolo Zunino This is my paper · ORCID
classification math.NAcs.NA MSC 65N5565N3068T07
keywords domaindecompositionmodelorderreductionneuralsurrogatemodelsRobin-Robinmethodmixed-dimensionalproblemslocalrepresentabilitymultiscalesimulation3D-1Dcoupling
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 proposes a way to solve multiscale problems without ever assembling or solving the global fine-scale system. It decomposes the domain into subdomains, maps every subdomain to a common reference shape, and learns three local maps — the local solution, the Robin transmission data, and the coarse residual contribution — with neural networks. These surrogates are inserted into a two-level Robin-Robin iteration, and the paper proves a stability result: if each local surrogate has uniform error epsilon, the global iteration converges to a neighborhood of the true solution of size at most CDD/(1-q) epsilon. The method is tested on a 3D-1D blood-perfusion model with unseen vascular geometries, where it reaches a stable error plateau without losing interface stability, and it weak-scales to over a million degrees of freedom.

What carries the argument

The load-bearing object is the local representability property: after mapping each subdomain to a common reference domain, all local solution maps that appear during the DD iteration are assumed to belong to one parametrized family s*(bµ) on a reference problem. This justifies training a single reusable surrogate. The iteration itself is a two-level non-overlapping Robin-Robin method; the paper interprets the surrogate-embedded version as a perturbed fixed-point iteration DD + δ, and Proposition 1 bounds the propagation of δ via the contraction factor q and a stability constant CDD. A cascaded training procedure (transmission first, then solution, then coarse residual) makes each surrogate's

What would settle it

Take a local Robin subproblem whose localized parameter lies outside the training distribution — e.g., a vascular graph with vessel radius or branching density substantially different from the 75 synthetic networks — and measure the surrogate's error supremum. If the measured global error plateau exceeds CDD/(1-q) ε (or the local surrogate errors exceed ε), the representability or uniform-accuracy premise is falsified.

Watch

Extended reading notes

Core claim

The central claim is that domain decomposition turns a globally intractable solution map into a family of locally representable operators, and that these local operators can be replaced by neural surrogates without destroying the convergence of the iterative solver. Concretely, for the mixed-dimensional 3D-1D problem, the paper builds three surrogates — the local solution map Srom, the Robin transmission map Trom, and the coarse-residual map Crom — deployed inside a two-level Robin-Robin iteration. Proposition 1 states that if the exact DD operator is a contraction with factor q and if the local surrogate error is uniformly bounded by epsilon, then the surrogate iteration satisfies limsup ||

Load-bearing premise

The entire error bound collapses if Assumptions 1 and 2 fail: every local subproblem that arises during the iteration must be exactly representable by one reference family on a common domain, and the finite training set must make the surrogate uniformly accurate (error ≤ ε) over the whole localized parameter space — neither premise is verified beyond the specific generative model used for training.

Editorial extensions

If this is right

  • Uniform local surrogate accuracy (error ≤ ε over the localized parameter space) is sufficient to keep the global iteration within a bounded error floor of size CDD/(1-q) ε.
  • The online stage needs no fine-scale operator assembly and no local high-fidelity solves: only neural evaluations and a small deterministic coarse solve remain.
  • The same reference surrogates can be reused across unseen microstructures and across larger decompositions without retraining, as demonstrated on 27, 64, and 125 subdomains.
  • The two-level coarse correction preserves the contraction behavior as the number of subdomains grows, so the surrogate iteration inherits weak scalability from the underlying DD method.
  • The cascaded training strategy reduces the offline-online distribution shift, so each surrogate is accurate on the perturbed states it actually encounters when deployed.

Reading between the lines

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

  • A direct corollary the paper does not develop: the error floor CDD/(1-q) ε can be used to allocate training effort — local samples should target the subdomains and DD states contributing the largest residual or interface errors, which the weak-scaling results already hint at.
  • The abstract formulation is not tied to the 3D-1D setting; if representability holds, the same operator-level replacement could apply to other parametrized multiscale PDEs with reusable local decompositions, but the assumption would have to be re-verified for each new problem class.
  • Because Srom does not enforce interface continuity in its outputs, the observed interface jumps (order 1e-3, with pointwise errors near 5%) suggest that adding an interface-consistency penalty to the training loss could lower the broken-energy plateau without changing the non-intrusive structure.
  • The reported online speedup of about 2.3x per subdomain excludes batched GPU inference; a parallel implementation that evaluates all subdomain surrogates as one batch would likely raise the speedup considerably.
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 MS-DD-ROM, a non-intrusive reduced-order framework for multiscale mixed-dimensional problems. Domain decomposition (DD) is used to localize the solution operator; three neural surrogates replace the local fine-scale operations: the transmission map, the local solution map, and the coarse-residual contribution. The paper formalizes local representability, analyzes the surrogate-perturbed DD iteration as an inexact contraction, and proves that the asymptotic global error is bounded by CDD/(1-q) times the uniform surrogate error (Proposition 1). A two-level Robin-Robin method for 3D-1D coupled problems is instantiated, with cascaded training designed to match online distributions. Numerical experiments on a 3D-1D oxygen-perfusion model show a stable error plateau on unseen microstructures and algorithmic weak scalability up to 125 subdomains.

Significance. The framework is timely and useful: it combines domain decomposition with learned local operators, avoids fine-scale assembly and local high-fidelity solves online, and proposes a cascaded training strategy to reduce offline-online distribution shift. The abstract perturbation analysis (Proposition 1) is correct under its assumptions, and the numerical experiments are extensive, including generalization to unseen geometries and weak scaling. The main value of the paper is conditional on verifying the stated uniform-accuracy assumption in the norm required by the theory; as it stands, the numerical evidence does not substantiate that assumption. The paper is a solid contribution to DD-ROMs and would be suitable for publication after the norm-mismatch and uniformity issues are addressed.

major comments (3)
  1. [§2.3, §4.2.1, Table 4] Assumption 2 requires sup_{bµ∈bP} ||e⋆(bµ)||_{V(Ω⋆)} ≤ ε, and Assumption 3 bounds the perturbation δ(k) in the broken-energy/interface norm ||·||⋆ of Proposition 2. However, all training losses in eqs. (15), (17), (18) and the reported errors in Table 4 are discrete L2(Ω⋆) or Euclidean for the coarse residual. No inequality is given linking these quantities to V(Ω⋆) or ||·||⋆. For Srom, whose closure term is microstructure-dependent, an L2-small error can be large in H1/energy near the embedded network. Therefore the reported 5.58% test error does not substantiate the ε entering Proposition 1. Please train/report errors in the energy/interface norm or provide and verify norm-equivalence constants over bP.
  2. [§5.2.1, Table 6] Assumption 2 is a supremum over the full localized parameter space bP, but the experimental coverage is limited to 65 training geometries, 5 DD iterations, and test geometries drawn from the same generative model. The weak-scaling study shows the error plateau growing with Nsub (relative L2 error from 2.67% for 27 subdomains to 5.79% for 125 subdomains; broken-energy error from 7.76% to 15.8%). This degradation indicates that larger decompositions produce local DD states progressively less represented in the offline distribution, which is precisely the premise whose failure would collapse the Proposition 1 bound. The paper should either provide evidence that the localized parameter space is covered in the norm of Assumption 2, or soften the claim and discuss adaptive enrichment as a necessary component.
  3. [Appendix, Proposition 2, eqs. (24)-(25)] The proof concludes q = sqrt(1 - θ_{k-1}/2) with θ_{k-1} = A/(A+B) ∈ (0,1]. As stated, this q depends on the iterate and can approach 1 as θ_{k-1} → 0, so no uniform constant q < 1 independent of k is established. Proposition 1 requires a fixed contraction constant q < 1. A uniform bound needs an additional argument, e.g., using the inverse/trace inequality to bound B from above by a multiple of A, or a different contraction estimate. As written, the proof of Proposition 2 is incomplete for the purpose of supporting Proposition 1.
minor comments (4)
  1. [Appendix, Prop. 2] The phrase 'the bilinear formal(·,·)' should read 'the bilinear form a_l(·,·)'.
  2. [References] Reference [14] contains a typo: 'heterognous' should be 'heterogeneous'.
  3. [Table 3] The notation 'MINN 0.3' is defined in Section 4.1, but a one-line reminder in the table caption would improve readability.
  4. [§3.2] The weak formulation of the local Robin problem is given for a pair of subdomains; it may help to state explicitly that the case of multiple interfaces follows by summing the corresponding interface terms.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: Prop. 1 is a standard perturbed-contraction bound with explicit assumptions; surrogate predictions are not fitted to the target result.

full rationale

The claimed central result (Prop. 1) is derived from two explicit assumptions: DD is a contraction with constant q in the norm ||·||_⋆, and the surrogate-induced perturbation satisfies ||δ^(k)||_⋆ ≤ C_DD ε. The appendix proof is precisely the standard recurrence ||e^(k+1)||_⋆ ≤ q||e^(k)||_⋆ + C_DD ε, giving limsup ≤ C_DD ε/(1−q). This is an a priori inequality, not an equation fitted to the observed error plateau. The numerical plateau in Section 5.2 and Table 5 is reported as consistent with the bound, and the bound is not used to define ε or C_DD from the measured errors. The surrogates are trained on 65 vascular geometries; the test geometries (Section 5.1) and the 64/125-subdomain weak-scaling cases (Section 5.2.1) are not in the training set, so the reported accuracies are genuine held-out evaluations. Self-citations to MINN [34] and POD-MINN+ [35] supply architecture choices, not the convergence statement; no uniqueness theorem is imported. The main caveats—Assumption 2 (sup over bP) is not verified in the energy/interface norm used in Prop. 1, since training losses (eqs. 15, 17, 18) are L2 or Euclidean, and the accuracy plateau degrades with Nsub—are verification/correctness gaps, not circular definitions or fitted predictions. Hence no significant circularity.

Assumptions & free parameters 5 free parameters · 7 assumptions · 0 invented entities

The framework does not postulate new physical entities, but it depends on several strong modeling assumptions: exact local representability after pullback, uniform surrogate accuracy over the full localized parameter space, and a stable/contractive exact DD iteration. The neural weights, POD basis, architecture radii, and Robin parameter are fitted or hand-chosen quantities that directly determine the reported accuracy.

free parameters (5)
  • Surrogate network weights θT, θS, θC = ~8.5M trainable parameters (Trom 1.39M, Srom 3.26M+1.91M, Crom 1.95M)
    Fitted by minimizing empirical losses on local DD data; the accuracy of the central claim depends directly on these fitted parameters.
  • POD basis V with 100 modes = Nrb = 100
    Data-derived linear subspace from high-fidelity snapshots; its projection error limits Srom accuracy.
  • Mesh-informed hyperparameters (interaction radii, auxiliary mesh size) = rS1=rS2=0.3, rSc=0.2; rT1=0.3, rT2=0.2, rT3=0.15; rC1=rC2=0.3
    Hand-chosen architectural constants that control receptive field and accuracy.
  • Robin parameter ρ = γ/h = γ ≥ ℓj; exact numerical value not reported
    Tuned to optimize DD convergence and required by the contraction proof in Proposition 2.
  • Cascaded loss weights ωrb, ωTS, ωST = not reported
    Appear in Eq. (16) for the Srom training objective; values are not given, so the exact training objective is not reproducible.
assumptions (7)
  • domain assumption Local representability (Assumption 1): every local subproblem equals a reference problem on a common domain after pullback.
    Foundation for reusing one surrogate across all subdomains and geometries; stated in Section 2.2 but not constructed or verified.
  • domain assumption Uniform approximation property (Assumption 2): sup_{bµ∈bP} ||e⋆(bµ)|| <= ε.
    Needed for the global error bound in Proposition 1; only finite-sample empirical errors are reported.
  • domain assumption Stability of DD with respect to local perturbations (Assumption 3).
    Lipschitz continuity of the exact DD operator under local solver perturbations; asserted with reference to optimized Schwarz arguments.
  • domain assumption Contraction of the discrete Robin-Robin iteration under the hypotheses of Proposition 2 (coercivity, continuity, stable lifting, γ ≥ ℓj).
    The proof in the Appendix establishes contraction only under these assumptions; they are plausible but not numerically verified in the experiments.
  • domain assumption Well-posedness and formal meaning of the 3D-1D Dirac-delta coupling model.
    The strong form uses δΛ; the paper relies on prior 3D-1D formulations [30-32] for the regularized trace/coupling.
  • domain assumption Uniform tensor-product Cartesian subdomains with affine equivalence to a reference mesh.
    Required for the common reference configuration and reused surrogates; limits current applicability to non-tensor-product geometries.
  • domain assumption Training and test vascular geometries are drawn from the same stochastic generative model.
    The 'unseen geometry' claim is about held-out samples from one distribution, not out-of-distribution generalization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multiscale Mixed-Dimensional Simulation via Domain Decomposition and Non-Intrusive Neural Model Order Reduction." pith.science (2026). https://pith.science/paper/B43WM74N

@misc{pith2026260715171,
  author       = {Pith},
  title        = {Pith review of: Multiscale Mixed-Dimensional Simulation via Domain Decomposition and Non-Intrusive Neural Model Order Reduction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/B43WM74N}},
  note         = {Machine review of arXiv:2607.15171}
}
read the original abstract

Many computational models arising in science and engineering exhibit a multiscale structure that makes the assembly or direct solution of the global problem computationally prohibitive. Domain Decomposition (DD) methods overcome this limitation by replacing the global problem with a sequence of coupled local problems, whose iterative solution reconstructs the global response. This work introduces a method in the family of Domain Decomposition Reduced Order Models (DD-ROMs), based on the observation that DD naturally localizes not only the solution operator but also its geometric and parametric dependence. The central idea is that DD transforms a globally intractable solution map into a family of locally representable operators learnable from affordable local data after identification with a common reference configuration, a concept that we formalize through the notion of local representability. Non-intrusive neural surrogates are then trained to approximate the fine-scale local operations and embedded into the iterative solver. The training algorithm is based on a cascaded strategy designed to match the distributions encountered by the deployed surrogate iteration. We interpret the resulting DD method as a perturbed fixed-point iteration and establish that the global error remains bounded by the surrogate approximation error. The framework is instantiated for mixed-dimensional elliptic problems coupling three-dimensional bulk domains with embedded one-dimensional inclusions, using a two-level non-overlapping Robin-Robin method. Numerical experiments show that the resulting DD-ROM is stable, achieves accurate approximation on unseen microscale geometries and features good scalability properties with respect to the number of subdomains, scaling to large size global problems while avoiding fine-scale operator assembly and local high-fidelity solvers in the online stage.

Figures

Figures reproduced from arXiv: 2607.15171 by the authors.

Figure 1
Figure 1. Schematic representation of the proposed framework. Domain decomposition localizes the global [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Cascaded training workflow. The transmission surrogate is trained on MS-DD-FOM trajectories; [PITH_FULL_IMAGE:figures/full_fig_p020_2.png] view at source ↗
Figure 3
Figure 3. Visualization of the pressure fields for the coupled 3D–1D perfusion problem. [PITH_FULL_IMAGE:figures/full_fig_p026_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Convergence curves of the MS-DD-FOM and MS-DD-ROM algorithms for the oxygen perfusion [PITH_FULL_IMAGE:figures/full_fig_p029_4.png]
Figure 5
Figure 5. Figure 5: Tissue pressure solution u ≡ p t on the domain Ω = [0 mm, 3 mm] 3 . From top to bottom, the rows correspond to slices taken in the upper, central, and lower regions of the domain along the z-direction. Left column: reference solution u ⋆ . Center column: MS-DD-ROM solu…
Figure 6
Figure 6. Figure 6: Weak-scaling study for the MS-DD-ROM algorithm applied to the oxygen perfusion problem [PITH_FULL_IMAGE:figures/full_fig_p031_6.png]
Figure 7
Figure 7. Figure 7: Average execution time per subdomain, reported in logarithmic scale, for the main computational [PITH_FULL_IMAGE:figures/full_fig_p033_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 2 linked inside Pith

  1. [1]

    D. W. Peaceman, Interpretation of well-block pressures in numerical reservoir simulation (includes associated paper 6988), Society of Petroleum Engineers Journal 18 (1978) 183–194

    1978

  2. [2]

    Jaffré, J

    J. Jaffré, J. Roberts, Modeling flow in porous media with fractures; discrete fracture models with matrix-fracture exchange, Numerical Analysis and Applications 5 (2012) 162–167

    2012

  3. [3]

    T. Koch, K. Heck, N. Schröder, H. Class, R. Helmig, A new simulation framework for soil–root interaction, evaporation, root growth, and solute transport, Vadose zone journal 17 (2018) 1–21

    2018

  4. [4]

    Schröder, M

    N. Schröder, M. Javaux, J. Vanderborght, B. Steffen, H. Vereecken, Effect of root water and solute uptake on apparent soil dispersivity: a simulation study, Vadose Zone Journal 11 (2012) vzj2012–0009

    2012

  5. [5]

    Possenti, S

    L. Possenti, S. di Gregorio, F. M. Gerosa, G. Raimondi, G. Casagrande, M. L. Costantino, P. Zunino, A computational model for microcirculation including fahraeus-lindqvist effect, plasma skimming and fluid exchange with the tissue interstitium, International Journal for Numerical Methods in Biomedical Engineering 35 (2019)

    2019

  6. [6]

    Possenti, A

    L. Possenti, A. Cicchetti, R. Rosati, D. Cerroni, M. Costantino, T. Rancati, P. Zunino, A mesoscale computational model for microvascular oxygen transfer, Annals of Biomedical Engineering 49 (2021) 3356–3373. 35

    2021

  7. [7]

    Vidotto, T

    E. Vidotto, T. Koch, T. Koppl, R. Helmig, B. Wohlmuth, Hybrid models for simulating blood flow in microvascular networks, Multiscale Modeling & Simulation 17 (2019) 1076–1102

    2019

  8. [8]

    Kremheller, A.-T

    J. Kremheller, A.-T. Vuong, B. A. Schrefler, W. A. Wall, An approach for vascular tumor growth based on a hybrid embedded/homogenized treatment of the vasculature within a multiphase porous medium model, International journal for numerical methods in biomedical engineering 35 (2019) e3253

    2019

Show all 42 references
  1. [9]

    Khristenko, S

    U. Khristenko, S. Schuß, M. Krüger, F. Schmidt, B. Wohlmuth, C. Hesch, Multidimensional coupling: A variationally consistent approach to fiber-reinforced materials, Computer Methods in Applied Mechanics and Engineering 382 (2021) 113869

    2021

  2. [10]

    Firmbach, A

    M. Firmbach, A. Popp, M. Mayr, I. Steinbrecher, Computational challenges in mixed- dimensional beam/solid coupling, in: Proceedings in Applied Mathematics & Mechanics, 2023

    2023

  3. [11]

    Edelvik, M

    F. Edelvik, M. Görtz, F. Hellman, G. Kettil, A. Målqvist, Numerical homogenization of spatial network models, Computer Methods in Applied Mechanics and Engineering 418 (2024) 116593

    2024

  4. [12]

    Bensoussan, J

    A. Bensoussan, J. Lions, G. Papanicolaou, Asymptotic Analysis for Periodic Structures, AMS Chelsea Publishing Series, American Mathematical Society, 2011

    2011

  5. [13]

    Efendiev, T

    Y. Efendiev, T. Y. Hou, Multiscale finite element methods: theory and applications, volume 4, Springer Science & Business Media, 2009

    2009

  6. [14]

    Weinan, B

    E. Weinan, B. Engquist, The heterognous multiscale methods, Communications in Mathematical Sciences 1 (2003) 87–132

    2003

  7. [15]

    Henning, A

    P. Henning, A. Målqvist, Localized orthogonal decomposition techniques for boundary value problems, SIAM Journal on Scientific Computing 36 (2014) A1609 – A1634. Cited by: 76; All Open Access, Green Open Access

    2014

  8. [16]

    Benner, S

    P. Benner, S. Gugercin, K. Willcox, A survey of projection-based model reduction methods for parametric dynamical systems, SIAM review 57 (2015) 483–531

    2015

  9. [17]

    Quarteroni, A

    A. Quarteroni, A. Manzoni, F. Negri, Reduced basis methods for partial differential equations: an introduction, volume 92, Springer, 2015

    2015

  10. [18]

    Quarteroni, P

    A. Quarteroni, P. Gervasio, F. Regazzoni, Combining physics-based and data-driven models: advancing the frontiers of research with scientific machine learning, arXiv preprint arXiv:2501.18708 (2025)

    2025 arXiv

  11. [19]

    Berner, P

    J. Berner, P. Grohs, G. Kutyniok, P. Petersen, The Modern Mathematics of Deep Learning, Cambridge University Press, 2022, p. 1–111

    2022

  12. [20]

    N. B. Kovachki, S. Lanthaler, A. M. Stuart, Operator learning: Algorithms and analysis, Handbook of Numerical Analysis 25 (2024) 419–467. 36

    2024

  13. [21]

    Boullé, A

    N. Boullé, A. Townsend, A mathematical guide to operator learning, in: Handbook of Numerical Analysis, volume 25, Elsevier, 2024, pp. 83–125

    2024

  14. [22]

    Quarteroni, Domain decomposition methods, in: Numerical models for differential problems, Springer, 2017, pp

    A. Quarteroni, Domain decomposition methods, in: Numerical models for differential problems, Springer, 2017, pp. 555–612

    2017

  15. [23]

    Toselli, O

    A. Toselli, O. Widlund, Domain Decomposition Methods: Algorithms and Theory, vol- ume 34 ofSpringer Series in Computational Mathematics, Springer, 2005

    2005

  16. [24]

    S. Ruan, A. G. Class, G. Rozza, A structured review of reduced order modeling for domain decomposition problems: State of the art and perspectives, Archives of Computational Methods in Engineering (2026)

    2026

  17. [25]

    Discacciati, J

    N. Discacciati, J. S. Hesthaven, Localized model order reduction and domain decomposition methods for coupled heterogeneous systems, International Journal for Numerical Methods in Engineering 124 (2023) 3964–3996

    2023

  18. [26]

    Taddei, X

    T. Taddei, X. Xu, L. Zhang, A non-overlapping optimization-based domain decomposi- tion approach to component-based model reduction of incompressible flows, Journal of Computational Physics 509 (2024) 113038

    2024

  19. [27]

    Farcas, R

    I.-G. Farcas, R. P. Gundevia, R. Munipalli, K. E. Willcox, Domain decomposition for data-driven reduced modeling of large-scale systems, AIAA Journal 62 (2024) 4071–4086

    2024

  20. [28]

    Heinlein, A

    A. Heinlein, A. Klawonn, M. Lanser, J. Weber, Combining machine learning and domain decomposition methods for the solution of partial differential equations—a review, GAMM- Mitteilungen 44 (2021) e202100001

    2021

  21. [29]

    M. J. Gander, F. Kwok, Best robin parameters for optimized schwarz methods at cross points, SIAM Journal on Scientific Computing 34 (2012) A1849–A1879

    2012

  22. [30]

    Laurino, P

    F. Laurino, P. Zunino, Derivation and analysis of coupled pdes on manifolds with high dimensionality gap arising from topological model reduction, ESAIM: Mathematical Modelling and Numerical Analysis 53 (2019) 2047 – 2080

    2019

  23. [31]

    Kuchta, F

    M. Kuchta, F. Laurino, K. A. Mardal, P. Zunino, Analysis and approximation of mixed- dimensional pdes on 3d-1d domains coupled with lagrange multipliers, SIAM Journal on Numerical Analysis 59 (2021) 558–582

    2021

  24. [32]

    Heltai, P

    L. Heltai, P. Zunino, Reduced lagrange multiplier approach for non-matching coupling of mixed-dimensional domains, Mathematical Models and Methods in Applied Sciences 33 (2023) 2425 – 2462

    2023

  25. [33]

    Klawonn, P

    A. Klawonn, P. Radtke, O. Rheinbach, A comparison of adaptive coarse spaces for iterative substructuring in two dimensions, Electron. Trans. Numer. Anal 45 (2016) 75–106

    2016

  26. [34]

    N. R. Franco, A. Manzoni, P. Zunino, Mesh-informed neural networks for operator learning in finite element spaces, Journal of Scientific Computing 97 (2023) 35

    2023

  27. [35]

    Vitullo, A

    P. Vitullo, A. Colombo, N. R. Franco, A. Manzoni, P. Zunino, Nonlinear model order reduction for problems with microstructure using mesh informed neural networks, Finite Elements in Analysis and Design 229 (2024) 104068. 37

    2024

  28. [36]

    Hesthaven, S

    J. Hesthaven, S. Ubbiali, Non-intrusive reduced order modeling of nonlinear problems using neural networks, Journal of Computational Physics 363 (2018) 55–78

    2018

  29. [37]

    Glorot, Y

    X. Glorot, Y. Bengio, Understanding the difficulty of training deep feedforward neural networks, Journal of Machine Learning Research - Proceedings Track 9 (2010) 249–256

    2010

  30. [38]

    Kutyniok, P

    G. Kutyniok, P. Petersen, M. Raslan, R. Schneider, A theoretical analysis of deep neural networks and parametric pdes, Constructive Approximation 55 (2022) 73–125

    2022

  31. [39]

    Paszke, Pytorch: An imperative style, high-performance deep learning library, arXiv preprint arXiv:1912.01703 (2019)

    A. Paszke, Pytorch: An imperative style, high-performance deep learning library, arXiv preprint arXiv:1912.01703 (2019)

    1912 arXiv

  32. [40]

    D. C. Liu, J. Nocedal, On the limited memory bfgs method for large scale optimization, Mathematical programming 45 (1989) 503–528

    1989

  33. [41]

    Raina, A

    R. Raina, A. Madhavan, A. Y. Ng, Large-scale deep unsupervised learning using graphics processors, in: Proceedings of the 26th annual international conference on machine learning, 2009, pp. 873–880

    2009

  34. [42]

    Assume thatu∈ Vis the unique fixed point ofDD, namely u=DD(u)

    Appendix Proposition1 (Stability under surrogate perturbations).Let( V,∥ · ∥⋆)be a normed space and letDD:V → Vbe a contraction with constantq∈[0,1), that is, ∥DD(v)− DD(w)∥⋆ ≤q∥v−w∥ ⋆ ∀v, w∈ V. Assume thatu∈ Vis the unique fixed point ofDD, namely u=DD(u). Let(eu(k))k≥0 be th...

Pith tools

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