Pith. sign in

REVIEW 4 major objections 6 minor 259 references

This review argues that Bayesian probability theory—Bayes' theorem plus the marginalization rule—is the single theoretical framework underlying forward and inverse uncertainty quantification in mechanics, unifying parameter calibration, sur

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 14:31 UTC pith:MWH55DEX

load-bearing objection A useful, mostly accurate review that reorganizes known Bayesian methods for mechanics; it has a few small but real defects that need fixing, and no new results beyond the authors' prior work. the 4 major comments →

arxiv 2607.18734 v1 pith:MWH55DEX submitted 2026-07-21 physics.comp-ph physics.data-anstat.MEstat.ML

Uncertainty quantification in mechanics: A unified Bayesian perspective

Sascha Ranftl , Malte Rolf , Gerhard A. Holzapfel , Ellen Kuhl This is my paper
classification physics.comp-ph physics.data-anstat.MEstat.ML
keywords Bayesian inferenceuncertainty quantificationinverse problemssurrogate modelingmodel calibrationmodel selectionexperimental designsensitivity analysis
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper tries to establish that uncertainty quantification in mechanics, usually split into forward and inverse problems, is not a collection of disparate methods but one coherent Bayesian calculus. It shows that both inverse tasks—calibrating constitutive parameters from experiments and training surrogate models from simulation data—are the same inference operation, differing only in what the likelihood targets. From there it derives model selection, experimental design, sensitivity analysis, and random fields for spatial heterogeneity as consequences of the same two rules: Bayes' theorem and the marginalization rule. A sympathetic reader cares because a single calculus lets practitioners carry uncertainty estimates, priors, and evidence-based model choice across every stage of a simulation workflow, particularly in biomechanics, where data are noisy and patient variability is large.

Core claim

The paper's central claim is that every task that UQ practitioners in mechanics perform—forward propagation, parameter calibration, surrogate construction, model selection, experimental design, sensitivity analysis, and spatial heterogeneity—is an instance of one probabilistic calculus built on Bayes' theorem and the marginalization rule. Surrogate training is reframed as a likelihood-based inverse problem; the forward uncertainty propagation integral is just marginalization; model selection compares evidence integrals; experimental design maximizes expected information gain; sensitivity indices are conditional variances; random fields are spatial priors. The review demonstrates the unificat

What carries the argument

The central object is the pair (Bayes' theorem, the marginalization rule), derived from logical consistency desiderata for partial belief. Bayes' theorem updates belief when new data arrive; the marginalization rule turns the joint distribution over inputs and outputs into the output distribution—the forward UQ integral—and also supplies the evidence normalization needed for model comparison. A second load-bearing device is the delta-likelihood assumption for deterministic simulations: when a simulation output is a unique function of its inputs, the general UQ equation collapses into the familiar expectation and variance integrals over the input distribution. This assumption is what lets the

Load-bearing premise

The strongest claim depends on the classical theorem that consistent partial belief is uniquely Bayesian; if that theorem's regularity assumptions fail, the unification becomes one workable framework rather than the unique one, and the practical workflow also assumes deterministic simulation outputs are uniquely determined by their inputs.

What would settle it

A single deterministic simulation that maps one input to multiple outputs—for example, cyclic loading with hysteresis where the same stress corresponds to different strain states—directly breaks the delta-likelihood reduction, a limitation the paper itself flags in a remark. For the uniqueness claim, a counterexample to the consistency desiderata—two equivalent inference paths yielding different probability assignments under a defensible real-valued, Boolean-limiting calculus—would settle whether Bayesian probability is genuinely the unique consistent calculus of partial belief.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Practitioners can switch between calibrating a constitutive model and training a surrogate without changing the underlying calculus: both are inverse problems defined by a likelihood and a prior.
  • Surrogate-based uncertainty quantification becomes fully Bayesian: the posterior over surrogate parameters propagates into output predictions through the marginalization rule, rather than relying only on point estimates.
  • Model selection and experimental design become the same Bayesian bookkeeping: compare model evidence integrals, then choose the next experiment to maximize expected information gain.
  • Sensitivity analysis is a special case of the framework: Sobol' indices are conditional variances, and a fully Bayesian version marginalizes surrogate-parameter uncertainty before computing those variances.
  • Random fields for heterogeneous materials become a standard prior modeling step, so spatial variability in tissue properties and constitutive parameter calibration fit under one coherent framework.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The paper leaves implicit a software-level consequence: a single probabilistic modeling language—priors, likelihoods, posterior samplers, and evidence estimators—could implement every UQ task in the review, replacing task-specific toolboxes.
  • The Bayesian Sobol' index defined by posterior-marginalized variances suggests a testable design rule: when surrogate training data are scarce, design decisions based on marginalized indices should differ from point-estimate indices, and the paper indicates the difference is quantifiable in closed form for polynomial chaos surrogates.
  • If the uniqueness theorem is accepted, the framework predicts that any alternative uncertainty calculus—interval analysis, fuzzy logic, or belief functions—must agree with Bayesian answers on every problem where its own consistency assumptions hold; that could be probed on mechanical calibration benchmarks.
  • The random-field-as-prior framing points to a practical extension: treating material microstructure statistics and constitutive parameter uncertainty as one hierarchical prior would let imaging data directly inform the covariance kernel of the spatial prior.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript is a review/position paper arguing that Bayesian probability theory provides a single, unified calculus for uncertainty quantification in mechanics. It frames forward uncertainty propagation as marginalization and treats model calibration, surrogate construction, model selection, experimental design, sensitivity analysis, and random-field priors as instances of Bayes' theorem plus the marginalization rule. The paper develops a consistent notation and workflow (Section 3), reviews PCE, Gaussian processes, neural networks/PINNs/neural operators and alternative surrogates (Section 4), derives BIC and an analytic PCE model evidence (Section 5), outlines Bayesian experimental design and optimization (Section 6), introduces Bayesian Sobol' indices (Section 7), and interprets random fields as spatial priors (Section 8). Worked examples include hierarchical inference of Young's modulus (Section 3.4), PCE-based arterial cylinder propagation (Section 4.8), and inverse design of a 3D-printed composite (Section 6.3).

Significance. If the framing is accepted, the paper offers a valuable pedagogical synthesis that connects a wide literature under one probabilistic calculus, which is useful for computational and biomechanics audiences. Its strengths include the explicit unified notation, the breadth of the review, the worked synthetic examples, and several candid caveats—notably the delta-likelihood remark (Section 3.1.2), the MAP/Laplace discussion (Section 3.4), and the 'incredulous' unknown-variance case (Section 5.4). The quasi-novel results, Eq. (127) and Eq. (165), are clearly attributed to the authors' prior work. However, the advertised uniqueness claim based on Cox's theorem is stronger than the assumptions stated, and the expected-utility formula in Eq. (139) contains a mathematical slip. These issues are local and fixable, but they affect the presentation of the central claims.

major comments (4)
  1. [Section 1 and Section 2, footnote 1] The opening sentence asserts that Bayesian probability theory 'has been mathematically proven as the unique, consistent approach for inference under uncertainty.' The three desiderata listed in footnote 1 are not sufficient for this uniqueness claim: Cox's theorem requires additional regularity, associativity, and continuity/monotonicity assumptions, and Halpern-style counterexamples satisfy qualitative consistency without additive probability. Since the advertised unification rests on this uniqueness sentence, I recommend softening it to 'a consistent approach' and either stating the extra assumptions or citing the counterexample literature. The practical workflow in Sections 3–8 does not depend on uniqueness and survives this change.
  2. [Section 6.1, Eq. (139)] Equation (139) is inconsistent with the correct nested expression in Eq. (137). The expected utility should be E[U] = ∫ U(y*) p(y*|D) dy*, with p(y*|D) = ∫ p(y*|D,θ) p(θ|D) dθ. As written, Eq. (139) inserts an extra ∫ p(θ) dθ and an additional dθ integration, making the formula measure-theoretically wrong and potentially misleading for implementation. Please correct the displayed equation and align it with Eq. (137).
  3. [Section 5.4, Eq. (127)] Equation (127) is called the 'model evidence' under a flat prior on PCE coefficients and Jeffreys' prior on the scale. A flat prior over an unbounded coefficient space is improper, so the marginal likelihood is defined only up to an arbitrary multiplicative constant. In nested models of different dimension, such an arbitrary constant does not cancel, making Bayes-factor comparisons non-unique. The data-independent factors Ω_{P_m} and the Gamma functions in Eq. (127) reflect the choice of Lebesgue measure in P_m dimensions rather than a genuine prior. The claim that Eq. (127) is 'exact' evidence, and the resulting Occam-factor demonstration, therefore needs either a proper prior specification or an explicit limiting construction to be fully justified.
  4. [Section 3.1.2 and Section 1] The core workflow uses the delta likelihood for deterministic simulations (Eq. (17) and the following Remark). The Remark correctly notes that this fails for deterministic but non-unique maps, e.g., hysteresis or path-dependent material behavior—cases explicitly within the paper's advertised scope of 'general mechanical problems' and biomechanics. Because the delta reduction underpins the forward UQ equations (Eqs. (18)–(19)) and the surrogate predictive equations (Eqs. (40)–(42)), the unified framework as presented applies only to single-valued maps. Please state this limitation in the introduction and Section 3.1.2 and indicate how the likelihood should be generalized for history-dependent responses.
minor comments (6)
  1. [Section 4.7] The sentence after Eq. (103) says the NLPD 'corresponds to the average negative log marginal likelihood (Eq. (74)) evaluated on the test set.' This is not accurate: Eq. (74) is the log marginal likelihood for the training data; the NLPD is the average negative log predictive density on test points. Please reword.
  2. [Section 7.2, Eq. (166)] The notation 'Bayes ← ← →' is nonstandard and should be replaced with a proper equality or arrow notation. Also clarify that p(θ) in Eq. (165) is the posterior p(θ|D), not the prior.
  3. [Section 5.4, Eq. (132)] The symbol '∝∼' is not defined. The comment that the log ratio 'scales as -log(N-P_m-2)' is imprecise: the factor is 2π/(N-P_m-2), so the statement should be made precise without asymptotic notation that could imply a different rate.
  4. [Section 3.1.3] The phrase 'As we learned [10]' is informal and ambiguous; replace with 'As discussed in [10].'
  5. [Figure 5 caption] Typos: 'Y oung' and 'represenative' should be 'Young' and 'representative'; in the same caption, 'detaile view' should be 'detailed view.'
  6. [Section 4.4.1] Typo: 'actitivation' should be 'activation.'

Circularity Check

0 steps flagged

No circular derivation: the Bayesian unification is a reorganization of standard Bayesian identities; the only notable weakness is an overbroad Cox-uniqueness claim, which is a correctness risk, not circularity.

full rationale

The paper's central claim is that Bayesian probability theory unifies forward UQ, inverse calibration, surrogate construction, model selection, experimental design, sensitivity analysis, and random-field priors. Inspecting the derivation chain, each advertised component is obtained by applying the standard rules Bayes' theorem and marginalization to a stated probabilistic model: model calibration in Eq. (23), surrogate training in Eqs. (32)-(42), model evidence in Eqs. (110)-(111), expected utility in Eqs. (136)-(139), Bayesian Sobol' indices in Eqs. (164)-(165), and random fields as priors in Section 8. These are standard, externally established Bayesian identities; no fitted parameter is renamed as a prediction, and no equation is shown to equal its own input merely by construction. The closed-form PCE evidence in Eq. (127) and the Bayesian Sobol' expressions in Section 7.2 are attributed to the authors' own prior work [17], which is self-citation, but these results are auxiliary examples rather than load-bearing premises for the unification claim: the unified workflow would survive even if those closed forms were absent. The Cox-theorem uniqueness sentence in Section 1 and footnote 1 does overstate what the three listed desiderata alone prove (Halpern-style counterexamples show that the uniqueness conclusion needs additional regularity assumptions), but this is an external mathematical accuracy issue, not circularity: the uniqueness claim is not justified by the authors' own prior work and is not used to derive the practical equations. Overall, no step in the paper's derivation chain reduces to its own inputs, so the circularity score is low.

Axiom & Free-Parameter Ledger

4 free parameters · 7 axioms · 0 invented entities

No new physical entities (particles, forces, dimensions) are postulated. The Bayesian Sobol' index (Eq. 165) is a new defined estimator, not an invented entity; it has no independent falsifiable handle outside the paper and is recommended mainly for its favorable statistical properties. All free parameters listed are example-level choices (Section 4.8 and 6.3), not parameters of the central claim: the review's thesis itself depends only on standard Bayesian axioms plus the domain assumptions above. The most consequential ledger entries are the improper-prior assumption in Section 5.4, which undermines the 'exact evidence' claim, and the BIC scaling assumption in Section 5.2, which is asserted so that the final form matches the standard criterion.

free parameters (4)
  • UCB exploration weight β = 0.8
    Hand-chosen in Section 6.3 to balance exploitation vs exploration in the inverse-design example; the reported trajectory of queries is sensitive to it.
  • GP hyperparameter priors (σ_f, σ, length scale) = HalfNormal(1); HalfNormal(0.3); LogNormal with median ≈ half the design range
    Weakly informative priors chosen by hand in Section 6.3 based on the standardized data scale; Fig. 9 additionally lists σ=0.35, which does not match the HalfNormal(0.3) noise prior stated in the text.
  • PCE order P and training sample count N = P=5, N=50 Latin-hypercube samples
    Hyperparameters chosen for the Section 4.8 arterial-cylinder example; the reported <0.001% surrogate error and the output distributions hold only for this synthetic setup.
  • Log-normal prior parameters for Young's modulus E = μ_lnE = 4.0, σ_lnE = 0.3
    Prescribed in Section 4.8 as coming from a previous Bayesian inverse problem; these values drive the propagated output posterior shown in Fig. 7.
axioms (7)
  • standard math Cox's theorem: a real-valued degree of belief that reduces to Boolean logic in the deterministic limit and is internally consistent must follow the sum and product rules of probability.
    Adopted in Sections 1-2 (footnote 1) as the uniqueness backbone of the claimed unification; known to require extra regularity assumptions to be fully rigorous (Halpern-style counterexamples).
  • domain assumption Additive Gaussian measurement noise, η ~ N(0, σ²).
    Eq. (15) and Eq. (22): the likelihood model for all worked examples; the paper acknowledges it is a default convenience rather than a derived fact.
  • domain assumption Measurement errors are independent and identically distributed across experiments and data points.
    Eq. (22) and Section 3.4: enables the factorization of the joint likelihood into a product.
  • ad hoc to paper Flat (improper) prior on PCE coefficients in the model-evidence derivation.
    Section 5.4, Eq. (127): required for the closed-form evidence; because the prior is improper, the marginal likelihood (and hence the claim that Eq. (127) is 'exact') is defined only up to an arbitrary normalization constant.
  • ad hoc to paper BIC scaling assumptions: prior width Λ_m ≈ σ and likelihood width Δ ~ σ/√N with an O(1) constant.
    Section 5.2, Eqs. (115)-(121): the O(1) proportionality constant and the Λ≈σ identification are asserted, not derived, so the prefactor collapses to P log N.
  • domain assumption Conditional independence and Markov structure in data assimilation.
    Eq. (25), Section 3.2.2: the Kalman-style factorization p(x_0:T | y_1:T) requires the Markov property for the state transition and observation models.
  • domain assumption Deterministic models admit a delta likelihood.
    Eq. (17), Section 3.1.2: exact only when the map x→y is unique; the paper's own Remark flags hysteresis and path-dependent materials as counterexamples.

pith-pipeline@v1.3.0-alltime-deepseek · 55793 in / 23406 out tokens · 183166 ms · 2026-08-01T14:31:26.351278+00:00 · methodology

0 comments
read the original abstract

Uncertainty quantification (UQ) is essential to experimental mechanics, but has become particularly relevant in computational mechanics, manifesting in two fundamental problem types: forward and inverse problems. The former addresses how input uncertainties propagate to the quantities of interest, whereas the latter aims to infer unknown parameters from experimental observations or simulations. Since efficient propagation typically requires a prohibitive number of evaluations to compute marginal output distributions, the development of fast, data-driven surrogate models becomes necessary. Thus, we can distinguish between two inverse tasks: (i) the identification and calibration of input uncertainties, and (ii) the construction of surrogates, a methodology collectively referred to as surrogate-based UQ. Building on probabilistic reasoning and the concept of partial belief, we demonstrate that Bayesian probability theory provides a unified theoretical framework for addressing both problem types. We further show that Bayesian inference allows for the seamless incorporation of essential subproblems, including model selection for identifying the most probable model specifications and experimental design for optimizing data collection by identifying experiments or simulations that maximize expected information gain about parameters, among others such as connections to sensitivity analysis or the use of special priors like random fields. While this theoretical framework is presented for general mechanical problems, particular emphasis is placed on biomechanics, where variability and uncertainty is especially pronounced due to inherent biological heterogeneity, patient-specific variability, and noisy data.

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

259 extracted references · 67 canonical work pages · 6 internal anchors

  1. [1]

    InternationalJournalforNumericalMethodsin Biomedical Engineering32(8), 02755 (2016) https://doi.org/10.1002/cnm.2755

    Eck, V.G., Donders, W.P., Sturdy, J., Feinberg, J., Delhaas, T., Hellevik, L.R., Huberts, W.: A guide to uncertainty quantification and sensi- tivity analysis for cardiovascular applications. InternationalJournalforNumericalMethodsin Biomedical Engineering32(8), 02755 (2016) https://doi.org/10.1002/cnm.2755

    work page doi:10.1002/cnm.2755 2016

  2. [2]

    American Journal of Physics 14(1), 1–13 (1946) https://doi.org/10.1119/1

    Cox, R.T.: Probability, frequency and reason- able expectation. American Journal of Physics 14(1), 1–13 (1946) https://doi.org/10.1119/1. 1990764

    doi:10.1119/1 1946

  3. [3]

    Oxford University Press

    Sivia, D., Skilling, J.: Data Analysis: A Bayesian Tutorial, (2006). Oxford University Press

    2006

  4. [4]

    Wollner, M.P., Rolf-Pissarczyk, M., Holzapfel, G.A.: A reparameterization-invariant Bayesian framework for uncertainty estimation and cali- brationofsimplematerials.ComputationalMe- chanics, 1–31 (2025) https://doi.org/10.1007/ s00466-024-02573-2

    2025

  5. [5]

    (eds.) Maximum En- tropy and Bayesian Methods

    Garrett, A.J.M.: Whence the laws of probability? In: Erickson, G.J., Rychert, J.T., Smith, C.R. (eds.) Maximum En- tropy and Bayesian Methods. Fundamental Theories of Physics, pp. 71–86 (1998). https://doi.org/10.1007/978-94-011-5028-6_6 . Springer

    work page doi:10.1007/978-94-011-5028-6_6 1998

  6. [6]

    https: //doi.org/10.1017/CBO9780511790423

    Jaynes, E.T.: Probability Theory, (2003). https: //doi.org/10.1017/CBO9780511790423 . Cam- bridge University Press

    doi:10.1017/cbo9780511790423 2003

  7. [7]

    Reviews of Modern Physics83(3), 943–999 (2011) https://doi.org/10.1103/ RevModPhys.83.943

    von Toussaint, U.: Bayesian inference in physics. Reviews of Modern Physics83(3), 943–999 (2011) https://doi.org/10.1103/ RevModPhys.83.943

    2011

  8. [8]

    von der Linden, W., Dose, V., von Toussaint, U.: Bayesian Probability Theory: Applications in the Physical Sciences, 1st edn. (2014). https: //doi.org/10.1017/CBO9781139565608 . Cam- bridge University Press

    doi:10.1017/cbo9781139565608 2014

  9. [9]

    Axioms1(1), 38–73 (2012) https://doi

    Knuth, K.H., Skilling, J.: Foundations of infer- ence. Axioms1(1), 38–73 (2012) https://doi. org/10.3390/axioms1010038

    doi:10.3390/axioms1010038 2012

  10. [10]

    In: Holzapfel, G.A., Rolf, M., Xu, X.Y

    Wollner, M.P., Rolf, M., Ranftl, S., Holzapfel, G.A.: Uncertainty quantification in biomechan- ics: probabilistic fundamentals. In: Holzapfel, G.A., Rolf, M., Xu, X.Y. (eds.) Model Valida- tion and Uncertainty Quantification in Biome- chanics: Sources and Methods of Uncertainty and Variability Analysis, (2026). Academic Press

    2026

  11. [11]

    Exper- imental Mechanics48, 381–402 (2008) https: //doi.org/10.1007/s11340-008-9148-y

    Avril, S., Bonnet, M., Bretelle, A.-S., Grédiac, M.,Hild,F.,Ienny,P.,Latourte,F.,Lemosse,D., Pagano,S.,Pagnacco,E.,Pierron,F.:Overview ofidentificationmethodsofmechanicalparam- eters based on full-field measurements. Exper- imental Mechanics48, 381–402 (2008) https: //doi.org/10.1007/s11340-008-9148-y

    doi:10.1007/s11340-008-9148-y 2008

  12. [12]

    PhD thesis, Universtity of Nottingham, UK (1997)

    Haylock, R.G.E.: Bayesian Inference About Outputs of Computationally Expensive Algo- rithms with Uncertainty on the Inputs. PhD thesis, Universtity of Nottingham, UK (1997)

    1997

  13. [13]

    In: Bernardo, J.M., 60 et al.(eds.) Bayesian Staistics 6: Proceedings of the Sixth Valencia International Meeting June 6-10, pp

    O’Hagan,A.,Kennedy,M.C.,Oakley,J.E.:Un- certainty analysis and other inference tools for complex computer codes. In: Bernardo, J.M., 60 et al.(eds.) Bayesian Staistics 6: Proceedings of the Sixth Valencia International Meeting June 6-10, pp. 503–524 (1999). https://doi.org/10. 1093/oso/9780198504856.003.0022

    arXiv 1999

  14. [14]

    Kennedy, M.C., O’Hagan, A.: Predicting the outputfromacomplexcomputercodewhenfast approximationsareavailable.Biometrika87(1), 1–13(2000)https://doi.org/10.1093/biomet/87. 1.1

    doi:10.1093/biomet/87 2000

  15. [15]

    Biometrika89(4), 769–784 (2002) https://doi.org/10.1093/biomet/89.4.769

    Oakley,J.,O’Hagan,A.:Bayesianinferencefor the uncertainty distribution of computer model outputs. Biometrika89(4), 769–784 (2002) https://doi.org/10.1093/biomet/89.4.769

    work page doi:10.1093/biomet/89.4.769 2002

  16. [16]

    O’Hagan, A.: Bayesian analysis of computer codeoutputs:atutorial.ReliabilityEngineering &SystemSafety91(10–11),1290–1300(2006) https://doi.org/10.1016/j.ress.2005.11.025

    work page doi:10.1016/j.ress.2005.11.025 2006

  17. [17]

    In: Physical Sciences Forum, vol

    Ranftl, S., von der Linden, W.: Bayesian surro- gate analysis and uncertainty propagation. In: Physical Sciences Forum, vol. 3, p. 6 (2021). https://doi.org/10.3390/psf2021003006 . Max- Ent 2021 Proceedings

    doi:10.3390/psf2021003006 2021

  18. [18]

    org/10.1103/PhysRevE.101.043308

    Fokina,D.,Muravleva,E.,Ovchinnikov,G.,Os- eledets,I.:Microstructuresynthesisusingstyle- basedgenerativeadversarialnetworks.Physical Review E101(4), 043308 (2020) https://doi. org/10.1103/PhysRevE.101.043308

    work page doi:10.1103/physreve.101.043308 2020

  19. [19]

    JOM73(1), 90–102 (2021) https://doi.org/10

    Hsu, T., Epting, W.K., Kim, H., Abernathy, H.W., Hackett, G.A., Rollett, A.D., Salvador, P.A., Holm, E.A.: Microstructure generation via generative adversarial network for hetero- geneous, topologically complex 3D materials. JOM73(1), 90–102 (2021) https://doi.org/10. 1007/s11837-020-04484-y

    2021

  20. [20]

    1038/s41598-023-27574-8

    Lambard, G., Yamazaki, K., Demura, M.: Gen- eration of highly realistic microstructural im- ages of alloys from limited data with a style- basedgenerativeadversarialnetwork.Scientific Reports13(1), 566 (2023) https://doi.org/10. 1038/s41598-023-27574-8

    2023

  21. [21]

    Medical Image Analysis, 103703 (2025) https: //doi.org/10.1016/j.media.2025.103703

    Feldman,P.,Fainstein,M.,Siless,V.,Delrieux, C., Iarussi, E.: Recursive variational autoen- codersfor3Dbloodvesselgenerativemodeling. Medical Image Analysis, 103703 (2025) https: //doi.org/10.1016/j.media.2025.103703

    arXiv 2025

  22. [22]

    Biomechanics and Modeling in Mechanobiol- ogy16(5), 1519–1533 (2017) https://doi.org/ 10.1007/s10237-017-0903-9

    Liang, L., Liu, M., Martin, C., Elefteri- ades, J.A., Sun, W.: A machine learning approach to investigate the relationship be- tween shape features and numerically pre- dicted risk of ascending aortic aneurysm. Biomechanics and Modeling in Mechanobiol- ogy16(5), 1519–1533 (2017) https://doi.org/ 10.1007/s10237-017-0903-9

    work page doi:10.1007/s10237-017-0903-9 2017

  23. [23]

    In: Holzapfel, G.A., Rolf, M., Xu, X.Y

    Verstraeten, S., Brüning, J., Goubergrits, L., Schievano, S., Capelli, C., Rizzo, A., Rolf, M., Holzapfel, G.A., Spanjaards, M., Huberts, W., SIMCorconsortium:Generationandvalidation of virtual patient cohorts forin silicoclinical trials: achievements from the SIMCor project. In: Holzapfel, G.A., Rolf, M., Xu, X.Y. (eds.) Model Validation and Uncertainty ...

    2026

  24. [24]

    Scientific Re- ports14(1), 5041 (2024) https://doi.org/10

    Lyu, X., Ren, X.: Microstructure reconstruc- tion of 2D/3D random materials via diffusion- based deep generative models. Scientific Re- ports14(1), 5041 (2024) https://doi.org/10. 1038/s41598-024-54861-9

    2024

  25. [25]

    Computers in Biology and Medicine 198 Part B, 111216 (2025) https://doi.org/10

    Colmenarez, J.A., Dong, P., Li, X., Gu, L.: Biomechanics-trained diffusion model predicts post-PCI plaque morphology and rupture from IVOCT: towards an AI digital twin for stent planning. Computers in Biology and Medicine 198 Part B, 111216 (2025) https://doi.org/10. 1016/j.compbiomed.2025.111216

    arXiv 2025

  26. [26]

    Computer Methods in Ap- plied Mechanics and Engineering461 Part A, 119171 (2026) https://doi.org/10.1016/j.cma

    Vahidullah, T., Kuhl, E.: Generative AI for material design: a mechanics perspective from burgers to matter. Computer Methods in Ap- plied Mechanics and Engineering461 Part A, 119171 (2026) https://doi.org/10.1016/j.cma. 2026.119171

    arXiv 2026

  27. [27]

    In: Proceedings of the IEEE/CVF In- ternational Conference on Computer Vision, pp

    Yang, G., Huang, X., Hao, Z., Liu, M.-Y., Be- longie, S., Hariharan, B.: Pointflow: 3D point cloud generation with continuous normalizing flows. In: Proceedings of the IEEE/CVF In- ternational Conference on Computer Vision, pp. 4541–4550 (2019). https://doi.org/10.1109/ ICCV.2019.00464 61

    arXiv 2019

  28. [28]

    Computer Meth- ods in Applied Mechanics and Engineering 401, 115594 (2022) https://doi.org/10.1016/j

    Ranftl, S., Rolf-Pissarczyk, M., Wolkerstor- fer, G., Pepe, A., Egger, J., von der Lin- den, W., Holzapfel, G.A.: Stochastic mod- eling of inhomogeneities in the aortic wall and uncertainty quantification using a bayesian encoder–decoder surrogate. Computer Meth- ods in Applied Mechanics and Engineering 401, 115594 (2022) https://doi.org/10.1016/j. cma.20...

    arXiv 2022

  29. [29]

    Holzapfel, G.A., Niestrawska, J.A., Ogden, R.W., Reinisch, A.J., Schriefl, A.J.: Modelling non-symmetric collagen fibre dispersion in ar- terialwalls.JournalofTheRoyalSocietyInter- face12(106), 20150188 (2015) https://doi.org/ 10.1098/rsif.2015.0188

    arXiv 2015

  30. [30]

    World Scientific

    Jammalamadaka, S.R., Sengupta, A.: Topics in CircularStatisticsvol.5,(2001).https://doi.org/ 10.1142/4031 . World Scientific

    work page doi:10.1142/4031 2001

  31. [31]

    The Physical Re- view106(4), 620–630 (1957) https: //doi.org/10.1103/PhysRev.106.620

    Jaynes, E.T.: Information theory and statistical mechanics. The Physical Re- view106(4), 620–630 (1957) https: //doi.org/10.1103/PhysRev.106.620

    doi:10.1103/physrev.106.620 1957

  32. [32]

    Clarendon Press

    Jeffreys, H.: Theory of Probability, (1939). Clarendon Press

    1939

  33. [33]

    https: //doi.org/10.1002/9781118014967.JohnWiley & Sons, Inc

    Kroese, D.P., Taimre, T., Botev, Z.I.: Hand- book of Monte Carlo Methods, (2011). https: //doi.org/10.1002/9781118014967.JohnWiley & Sons, Inc

    work page doi:10.1002/9781118014967.johnwiley 2011

  34. [34]

    In: Holzapfel, G.A., Rolf, M., Xu, X.Y

    Pensalfini, M., Böl, M., Buganza Tepole, A.: Identification of model parameter distribu- tions via multilevel Bayesian approaches. In: Holzapfel, G.A., Rolf, M., Xu, X.Y. (eds.) Model Validation and Uncertainty Quantifica- tion in Biomechanics: Sources and Methods of Uncertainty and Variability Analysis, (2026). Academic Press

    2026

  35. [35]

    International Journal for Numeri- cal Methods in Biomedical Engineering, 3576 (2022) https://doi.org/10.1002/cnm.3576

    Ranftl,S.,Müller,T.S.,Windberger,U.,Brenn, G.,vonderLinden,W.:ABayesianapproachto blood rheological uncertainties in aortic hemo- dynamics. International Journal for Numeri- cal Methods in Biomedical Engineering, 3576 (2022) https://doi.org/10.1002/cnm.3576

    work page doi:10.1002/cnm.3576 2022

  36. [36]

    Journal of the Mechanical Behavior of Biomedical Materials59, 108–127 (2016) https://doi.org/10.1016/j.jmbbm.2015.10.025

    Madireddy, S., Sista, B., Vemaganti, K.: Bayesian calibration of hyperelas- tic constitutive models of soft tissue. Journal of the Mechanical Behavior of Biomedical Materials59, 108–127 (2016) https://doi.org/10.1016/j.jmbbm.2015.10.025

    work page doi:10.1016/j.jmbbm.2015.10.025 2016

  37. [37]

    Computer Meth- ods in Applied Mechanics and Engineering 357, 112604 (2019) https://doi.org/10.1016/j

    Teferra, K., Brewick, P.T.: A Bayesian model calibration framework to evaluate brain tissue characterization experiments. Computer Meth- ods in Applied Mechanics and Engineering 357, 112604 (2019) https://doi.org/10.1016/j. cma.2019.112604

    arXiv 2019

  38. [38]

    In: Sommer, G., Li, K., Haspinger, D.C., Ogden, R.W

    Sundnes, J., Rodríguez-Cantano, R.: A Bayesian approach to parameter estimation in cardiac mechanics. In: Sommer, G., Li, K., Haspinger, D.C., Ogden, R.W. (eds.) Solid (Bio)mechanics: Challenges of the Next Decade. Studies in Mechanobiology, Tissue Engineering and Biomaterials vol. 24, (2022). https://doi.org/10.1007/978-3-030-92339-6_ 10 . Springer Cham

    work page doi:10.1007/978-3-030-92339-6_ 2022

  39. [39]

    Computer Methods in Applied Mechanics and Engineer- ing398,115225(2022)https://doi.org/10.1016/ j.cma.2022.115225

    Joshi, A., Thakolkaran, P., Zheng, Y., Escande, M., Flaschel, M., De Lorenzis, L., Kumar, S.: Bayesian-EUCLID: discovering hyperelas- tic material laws with uncertainties. Computer Methods in Applied Mechanics and Engineer- ing398,115225(2022)https://doi.org/10.1016/ j.cma.2022.115225

    arXiv 2022

  40. [40]

    Krijnen, R.P., Joshi, A., Kumar, S., Peirlinck, M.: Unsupervised full-field Bayesian inference of orthotropic hyperelasticity from a single bi- axialtest:amyocardialcasestudy.arXiv(2025) https://doi.org/10.48550/arXiv.2510.09498

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2510.09498 2025

  41. [41]

    Philosophical Transactions of the Royal Society A: Mathematical, Physi- calandEngineeringSciences383(2292)(2025) https://doi.org/10.1098/rsta.2024.0223

    Richter, J., Nitzler, J., Pegolotti, L., Menon, K., Biehler,J.,Wall,W.A.,Schiavazzi,D.E.,Mars- den, A.L., Pfaller, M.R.: Bayesian Windkessel calibration using optimized zero-dimensional surrogate models. Philosophical Transactions of the Royal Society A: Mathematical, Physi- calandEngineeringSciences383(2292)(2025) https://doi.org/10.1098/rsta.2024.0223

    arXiv 2025

  42. [42]

    arXiv (2026) https://doi

    Choi, C.H., Marsden, A.L., Schiavazzi, D.E.: FalconBC: Flow matching for Amortized in- ference of Latent-CONditioned physiologic Boundary Conditions. arXiv (2026) https://doi. org/10.48550/arXiv.2603.19331 62

    work page doi:10.48550/arxiv.2603.19331 2026

  43. [43]

    Journal of Computational Physics 511, 113117 (2024) https://doi.org/10.1016/j

    Karumuri, S., Bilionis, I.: Learning to solve Bayesian inverse problems: an amortized vari- ational inference approach using Gaussian and flow guides. Journal of Computational Physics 511, 113117 (2024) https://doi.org/10.1016/j. jcp.2024.113117

    arXiv 2024

  44. [44]

    Composites Science and Technology228,109630(2022)https://doi.org/ 10.1016/j.compscitech.2022.109630

    Thomas, A.J., Barocio, E., Bilionis, I., Pipes, R.B.: Bayesian inference of fiber orientation andpolymerpropertiesinshortfiber-reinforced polymer composites. Composites Science and Technology228,109630(2022)https://doi.org/ 10.1016/j.compscitech.2022.109630

    arXiv 2022

  45. [45]

    Probabilistic Engineering Mechanics 66, 103153 (2021) https://doi.org/10.1016/j

    Chakraborty, A., Messner, M.C.: Bayesian analysis for estimating statistical parameter distributions of elasto-viscoplastic material models. Probabilistic Engineering Mechanics 66, 103153 (2021) https://doi.org/10.1016/j. probengmech.2021.103153

    arXiv 2021

  46. [46]

    Computational Mechanics66, 827–849 (2020) https://doi.org/ 10.1007/s00466-020-01876-4

    Khodadadian, A., Noii, N., Parvizi, M., Ab- baszadeh, M., Wick, T., Heitzinger, C.: A Bayesian estimation method for variational phase-field fracture problems. Computational Mechanics66, 827–849 (2020) https://doi.org/ 10.1007/s00466-020-01876-4

    work page doi:10.1007/s00466-020-01876-4 2020

  47. [47]

    Ad- vanced Modeling and Simulation in Engineer- ing Sciences9, 24 (2022) https://doi.org/10

    Willmann, H., Nitzler, J., Brandstaeter, S., Wall, W.A.: Bayesian calibration of coupled computational mechanics models under un- certainty based on interface deformation. Ad- vanced Modeling and Simulation in Engineer- ing Sciences9, 24 (2022) https://doi.org/10. 1186/s40323-022-00237-5

    2022

  48. [48]

    1016/j.jcp.2020.109913

    Yang, L., Meng, X., Karniadakis, G.E.: B- pinns: Bayesian physics-informed neural net- works for forward and inverse PDE prob- lemswithnoisydata.JournalofComputational Physics425, 109913 (2021) https://doi.org/10. 1016/j.jcp.2020.109913

    arXiv 2021

  49. [49]

    Computer Methods in Applied Mechanics and Engineering402, 115346 (2022) https://doi.org/10.1016/j.cma

    Linka, K., Schäfer, A., Meng, X., Zou, Z., Karniadakis, G.E., Kuhl, E.: Bayesian physics informed neural networks for real-world non- linear dynamical systems. Computer Methods in Applied Mechanics and Engineering402, 115346 (2022) https://doi.org/10.1016/j.cma. 2022.115346

    arXiv 2022

  50. [50]

    Computer Methods in Applied Mechanics and Engineering433, 117517 (2025) https://doi.org/10.1016/j.cma

    Linka, K., Holzapfel, G.A., Kuhl, E.: Dis- covering uncertainty: Bayesian constitutive ar- tificial neural networks. Computer Methods in Applied Mechanics and Engineering433, 117517 (2025) https://doi.org/10.1016/j.cma. 2024.117517

    arXiv 2025

  51. [51]

    Computer Methods and Programs in Biomedicine200, 105828 (2021) https://doi

    Xie, H., Song, J., Zhong, Y., Li, J., Gu, C., Choi, K.-S.: Extended Kalman filter nonlinear finite element method for nonlinear soft tissue deformation. Computer Methods and Programs in Biomedicine200, 105828 (2021) https://doi. org/10.1016/j.cmpb.2020.105828

    arXiv 2021

  52. [52]

    Journal of the Mechanical Behavior of Biomedical Materials 137, 105553 (2023) https://doi.org/10.1016/j

    Song,J.,Xie,H.,Zhong,Y.,Gu,C.,Choi,K.-S.: Maximum likelihood-based extended Kalman filter for soft tissue modelling. Journal of the Mechanical Behavior of Biomedical Materials 137, 105553 (2023) https://doi.org/10.1016/j. jmbbm.2022.105553

    arXiv 2023

  53. [53]

    Zhu, X., Li, J., Zhong, Y., Choi, K.-S., Shirin- zadeh, B., Smith, J., Gu, C.: Iterative Kalman filterforbiologicaltissueidentification.Interna- tional Journal of Robust and Nonlinear Control 35, 3949–3961 (2023) https://doi.org/10.1002/ rnc.6742

    2023

  54. [54]

    Journal of the Royal StatisticalSociety:SeriesB(StatisticalMethod- ology)63(3), 425–464 (2001) https://doi.org/ 10.1111/1467-9868.00294

    Kennedy,M.C.,O’Hagan,A.:Bayesiancalibra- tion of computer models. Journal of the Royal StatisticalSociety:SeriesB(StatisticalMethod- ology)63(3), 425–464 (2001) https://doi.org/ 10.1111/1467-9868.00294

    arXiv 2001

  55. [55]

    SIAM/ASA Journal on Un- certainty Quantification6(2), 457–496 (2018) https://doi.org/10.1137/16M1106419

    Morrison,R.E.,Oliver,T.A.,Moser,R.D.:Rep- resenting model inadequacy: a stochastic op- erator approach. SIAM/ASA Journal on Un- certainty Quantification6(2), 457–496 (2018) https://doi.org/10.1137/16M1106419

    work page doi:10.1137/16m1106419 2018

  56. [56]

    Journal of the Royal Society Interface 17(173), 20200886 (2020) https://doi.org/10

    Paun, L.M., Colebank, M.J., Olufsen, M.S., Hill, N.A., Husmeier, D.: Assessing model mismatch and model selection in a Bayesian uncertainty quantification analysis of a fluid- dynamics model of pulmonary blood circula- tion. Journal of the Royal Society Interface 17(173), 20200886 (2020) https://doi.org/10. 1098/rsif.2020.0886

    arXiv 2020

  57. [57]

    International 63 Journal for Numerical Methods in Biomedical Engineering38,3575(2022)https://doi.org/10

    Römer, U., Liu, J., Böl, M.: Surrogate-based Bayesian calibration of biomechanical models with isotropic material behavior. International 63 Journal for Numerical Methods in Biomedical Engineering38,3575(2022)https://doi.org/10. 1002/cnm.3575

    2022

  58. [58]

    In- verse Problems34, 025008 (2018) https://doi.org/10.1088/1361-6420/aaa34d

    Calvetti, D., Dunlop, M., Somersalo, E., Stuart, A.: Iterative updating of model error for Bayesian inversion. In- verse Problems34, 025008 (2018) https://doi.org/10.1088/1361-6420/aaa34d

    doi:10.1088/1361-6420/aaa34d 2018

  59. [59]

    PLoS Computational Biology19(3), 1010902 (2023) https://doi.org/ 10.1371/journal.pcbi.1010902

    Pensalfini, M., Buganza Tepole, A.: Mechano- biological and bio-mechanical pathways in cu- taneous wound healing. PLoS Computational Biology19(3), 1010902 (2023) https://doi.org/ 10.1371/journal.pcbi.1010902

    work page doi:10.1371/journal.pcbi.1010902 2023

  60. [60]

    SIAM Journal on Scientific Comput- ing31(5),3274–3300(2009)https://doi.org/10

    Koutsourelakis, P.-S.: Accurate uncertainty quantification using inaccurate computational models. SIAM Journal on Scientific Comput- ing31(5),3274–3300(2009)https://doi.org/10. 1137/080733565

    2009

  61. [61]

    Wiener,N.:Thehomogeneouschaos.American Journal of Mathematics60(4), 897–936 (1938) https://doi.org/10.2307/2371268

    doi:10.2307/2371268 1938

  62. [62]

    1137/S1064827501387826

    Xiu, D., Karniadakis, G.E.: The Wiener–Askey polynomial chaos for stochastic differential equations.SIAMJournalonScientificComput- ing24(2), 619–644 (2002) https://doi.org/10. 1137/S1064827501387826

    2002

  63. [63]

    Springer New York, NY

    Ghanem, R.G., Spanos, P.D.: Stochastic Finite Elements:ASpectralApproach,(1991).https:// doi.org/10.1007/978-1-4612-3094-6 . Springer New York, NY

    doi:10.1007/978-1-4612-3094-6 1991

  64. [64]

    (eds.): Handbook of Uncertainty Quantification vol

    Ghanem, R., Higdon, D., Owhadi, H. (eds.): Handbook of Uncertainty Quantification vol. 6, (2017). https://doi.org/10.1007/ 978-3-319-12385-1 . Springer Cham

    2017

  65. [65]

    http://tonyohagan.co.uk/academic/pdf/ Polynomial-chaos.pdf,accessed25.06.2019

    O’Hagan, A.: Polynomial chaos: a tutorial and critique from a statistician’s perspective (2013). http://tonyohagan.co.uk/academic/pdf/ Polynomial-chaos.pdf,accessed25.06.2019

    2013

  66. [66]

    Reliability Engineering & System Safety106, 179–190 (2012) https: //doi.org/10.1016/j.ress.2012.05.002

    Oladyshkin, S., Nowak, W.: Data-driven uncer- taintyquantificationusingthearbitrarypolyno- mial chaos expansion. Reliability Engineering & System Safety106, 179–190 (2012) https: //doi.org/10.1016/j.ress.2012.05.002

    doi:10.1016/j.ress.2012.05.002 2012

  67. [67]

    Com- puterMethodsinAppliedMechanicsandEngi- neering351,643–666(2019)https://doi.org/10

    Jakeman, J.D., Franzelin, F., Narayan, A., El- dred, M., Plfüger, D.: Polynomial chaos expan- sions for dependent random variables. Com- puterMethodsinAppliedMechanicsandEngi- neering351,643–666(2019)https://doi.org/10. 1016/j.cma.2019.03.049

    2019

  68. [68]

    Probabilistic Engineering Mechanics55, 1–16 (2019) https: //doi.org/10.1016/j.probengmech.2018.08.001

    Torre, E., Marelli, S., Embrechts, P., Sudret, B.: A general framework for data-driven un- certainty quantification under complex input dependencies using vine copulas. Probabilistic Engineering Mechanics55, 1–16 (2019) https: //doi.org/10.1016/j.probengmech.2018.08.001

    work page doi:10.1016/j.probengmech.2018.08.001 2019

  69. [69]

    Xiu, D., Karniadakis, G.E.: The Wiener-Askey polynomial chaos for stochastic differential equations.SIAMJournalonScientificComput- ing27(3), 1118–1139 (2005)

    2005

  70. [70]

    Reliability Engi- neering&SystemSafety93(7),964–979(2008) https://doi.org/10.1016/j.ress.2007.04.002

    Sudret, B.: Global sensitivity analysis using polynomial chaos expansions. Reliability Engi- neering&SystemSafety93(7),964–979(2008) https://doi.org/10.1016/j.ress.2007.04.002

    doi:10.1016/j.ress.2007.04.002 2008

  71. [71]

    Reliability Engineering & System Safety94(7),1161–1172(2009)https://doi.org/ 10.1016/j.ress.2008.10.008

    Crestaux, T., Le Maître, O.P., Martinez, J.- M.: Polynomial chaos expansion for sensitiv- ity analysis. Reliability Engineering & System Safety94(7),1161–1172(2009)https://doi.org/ 10.1016/j.ress.2008.10.008

    doi:10.1016/j.ress.2008.10.008 2009

  72. [72]

    IEEE Con- trolSystemsLetters2(1),169–174(2017)https: //doi.org/10.1109/LCSYS.2017.2778138

    Mühlpfordt, T., Findeisen, R., Hagenmeyer, V., Faulwasser, T.: Comments on truncation errors for polynomial chaos expansions. IEEE Con- trolSystemsLetters2(1),169–174(2017)https: //doi.org/10.1109/LCSYS.2017.2778138

    arXiv 2017

  73. [73]

    In: Hesthaven, J., Rønquist, E

    Zhang, Z., Choi, M., Karniadakis, G.E.: An- chor points matter in ANOVA decomposition. In: Hesthaven, J., Rønquist, E. (eds.) Spec- tral and High Order Methods for Partial Dif- ferential Equations. Lecture Notes in Com- putational Science and Engineering vol. 76, pp. 347–355 (2011). https://doi.org/10.1007/ 978-3-642-15337-2_32 . Springer Berlin

    2011

  74. [74]

    Journal of Computa- tional Physics229(5), 1536–1557 (2010) https://doi.org/10.1016/j.jcp.2009.10.043 64

    Foo, J., Karniadakis, G.E.: Multi-element probabilistic collocation method in high dimensions. Journal of Computa- tional Physics229(5), 1536–1557 (2010) https://doi.org/10.1016/j.jcp.2009.10.043 64

    work page doi:10.1016/j.jcp.2009.10.043 2010

  75. [75]

    Journal of Computational Physics 230(6), 2345–2367 (2011)

    Blatman, G., Sudret, B.: Adaptive sparse poly- nomial chaos expansion based on least angle regression. Journal of Computational Physics 230(6), 2345–2367 (2011)

    2011

  76. [76]

    Journal of Computational Physics 230(8), 3015–3034 (2011) https://doi.org/10

    Doostan, A., Owhadi, H.: A non-adapted sparse approximation of PDEs with stochas- tic inputs. Journal of Computational Physics 230(8), 3015–3034 (2011) https://doi.org/10. 1016/j.jcp.2011.01.002

    2011

  77. [77]

    org/10.1137/20M1315774

    Lüthen,N.,Marelli,S.,Sudret,B.:Sparsepoly- nomial chaos expansions: literature survey and benchmark.SIAM/ASAJournalonUncertainty Quantification9(2),593–649(2021)https://doi. org/10.1137/20M1315774

    work page doi:10.1137/20m1315774 2021

  78. [78]

    Journal of Computational Physics397, 108850 (2019) https://doi.org/10.1016/j.jcp.2019.07.048

    Zhang, D., Lu, L., Guo, L., Karniadakis, G.E.: Quantifying total uncertainty in physics- informed neural networks for solving forward and inverse stochastic problems. Journal of Computational Physics397, 108850 (2019) https://doi.org/10.1016/j.jcp.2019.07.048

    work page doi:10.1016/j.jcp.2019.07.048 2019

  79. [79]

    Analysis and Applications17(1) (2019) https://doi.org/10.1142/S0219530518500203

    Schwab, C., Zech, J.: Deep learning in high di- mension: neural network expression rates for generalized polynomial chaos expansions in UQ. Analysis and Applications17(1) (2019) https://doi.org/10.1142/S0219530518500203

    doi:10.1142/s0219530518500203 2019

  80. [80]

    Cooper,R.G.:AugmentedNeuralNetworkSur- rogate Models for Polynomial Chaos Expan- sionsandReducedOrderModeling.PhDthesis, VirginiaPolytechnicInstituteandStateUniver- sity, CA, USA (2021)

    2021

Showing first 80 references.