Pith. sign in

REVIEW 5 major objections 6 minor 43 references

Should We Simultaneously Calibrate Multiple Computer Models?

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Simultaneous calibration of multiple computer models is feasible and improves emulation accuracy, but it makes calibration parameters non-identifiable in high-dimensional problems.

desk verdict A useful, honest extension of neural data fusion that improves multi-fidelity emulation, but the calibration gains are not isolated and high-dimensional inference fails; the title question gets only a partial answer. read the letter →

arxiv 2505.18176 v2 pith:SYZ3672M submitted 2025-05-14 cs.LG stat.ML

Jonathan Tammer Eweis-Labolle , Tyler Johnson , Xiangyu Sun , Ramin Bostanabad This is my paper · ORCID
classification cs.LGstat.ML
keywords ModelCalibrationMulti-fidelityModelingUncertaintyQuantificationProbabilisticNeuralNetworksInverseProblemsManifoldLearningDataFusion
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

Computer models used in engineering often come in several versions with different fidelities, costs, and tuning parameters, and the usual practice is to calibrate each one separately against high-fidelity data. This paper asks whether calibrating all of them at once is better, and it builds a neural-network framework, iPro-NC, to find out. The framework learns a separate probability distribution for every calibration parameter of every model while using a shared latent embedding of the data sources, so it works when models have different numbers of parameters and different outputs. The paper's answer is mixed: simultaneous calibration improves emulation of the high-fidelity source in both test problems, including an engineering case, but in higher-dimensional problems the estimated parameters become non-identifiable and can sit at the edges of their sampling ranges. The authors conclude that the approach is worth using when accurate emulation is the goal, and that domain knowledge must be added when physical parameters matter.

What carries the argument

The load-bearing object is iPro-NC, a multi-block feed-forward network that converts multi-fidelity calibration into a latent-variable problem. A source-indicator variable $t_s$ is one-hot encoded and mapped by Block 0 to a source embedding $z_s$; masked calibration inputs are combined with $z_s$ in Block 1 to form $z_{\theta}$; categorical inputs go through Block 2; and Block 3 outputs Gaussian means and standard deviations for every response. Each calibration parameter of each LF source has its own learned mean and standard deviation, sampled through the reparameterization trick and clamped to the training domain. The loss combines negative log-likelihood and interval-score terms for both emulation and calibration, plus a KL term on the calibration standard deviations, with per-source normalization so unbalanced data does not dominate. Block 0's embedding is the interpretability device: distances in $z_s$ are meant to show which sources are most similar to the HF source.

What would settle it

Take a synthetic problem with a known true parameter vector $\theta^{\ast}$ and two LF sources, one that contains the HF model exactly at $\theta^{\ast}$ and one with a deliberately misspecified form. If training on both sources moves the posterior mode of the exact source's parameters away from $\theta^{\ast}$ while emulation error stays low, the benefit of simultaneity is coming from bias correction, not from better parameter identification. The paper's own analytic example already shows Block 0 can place the misspecified source closer to $s_0$ than the exact one, so re-running that example with the true $\theta^{\ast}$ recorded would settle whether the failure is inherent or fixable.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that multi-source calibration is feasible and can be strictly better than one-at-a-time calibration, provided the method is built to handle sources with different parameter spaces and outputs. iPro-NC encodes each source into a low-dimensional latent variable, routes calibration parameters through the network, and trains with a loss that separates emulation (reproduce each source's data) from calibration (make each low-fidelity source reproduce the high-fidelity data). In the analytic example, training on all sources made the HF emulator never the worst on any output, and in the engineering problem all-source training reduced RRMSE substantially for both LF-to-HF emulation and calibrated-LF predictions. The same flexibility, however, lets the network consume misspecification: with model-form error present, the source embedding can rank sources incorrectly, and in the high-dimensional hardening-law calibration most parameters are estimated at the extremes of their ranges. So the paper establishes both halves of the tradeoff: simultaneous calibration helps prediction, and it worsens parameter identifiability as input and parameter dimensionality grow.

Load-bearing premise

The network's latent representation can capture the true relationship between every low-fidelity source and the high-fidelity source, and the loss can tell apart errors that should be absorbed by calibration parameters from errors that should be absorbed by bias correction.

Editorial extensions

If this is right

  • When all available sources are used, iPro-NC's HF emulator is never the worst on any output in the analytic example, and all-source training cuts RRMSE in the engineering example by roughly an order of magnitude for calibrated LF predictions.
  • The framework removes the usual requirement that all models share the same calibration parameters: sources with different parameter counts and different responses are handled by masking and per-source distributions.
  • For low-dimensional problems, posterior distributions for calibration parameters cover the MSE-optimal values when the LF model has no model-form error, so joint posterior inference is meaningful there.
  • In higher-dimensional engineering problems the estimated parameters frequently land at the bounds of their sampling ranges, so parameter values from simultaneous calibration should not be read as physical quantities without added constraints.
  • The authors state directly that using iPro-NC is justified when maximizing emulation accuracy is the only goal, since training cost grows with the number of sources but calibration reliability does not.

Reading between the lines

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

  • A practical design rule follows that the paper leaves implicit: include extra LF sources when the deliverable is a surrogate or prediction, and restrict to a minimal, well-posed set of models when the deliverable is the parameter posterior itself.
  • The Block 0 embedding could be repurposed as a model-form-error diagnostic, but only if identifiability is enforced; otherwise the paper's analytic result shows the embedding can rank a misspecified source closer to the truth than the correct source.
  • Adding a physics-based penalty that encourages the source embedding to respect known fidelity ordering, or fixing a subset of physically meaningful parameters as in the two-step engineering calibration, is a direct testable extension that may recover identifiability in the high-dimensional regime.
  • The per-source normalized negative log-likelihood and the emulation/calibration split in the loss are general ideas that could transfer to other multi-fidelity surrogates beyond this architecture.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper asks whether multiple low-fidelity computer models should be calibrated simultaneously against a single high-fidelity source, and proposes iPro-NC, a multi-block neural-network framework that combines multi-fidelity emulation with per-source probabilistic calibration. The network encodes a source indicator and categorical variables into latent spaces, maps calibration parameters through a masked block, and outputs per-output normal distributions; the loss in Eq. (2) mixes emulation and calibration negative log-likelihood terms, interval scores, and a KL regularization term. The method is tested on an analytic three-source problem with known functional forms and on a finite-element tensile specimen problem with a Holloman high-fidelity model and two Voce low-fidelity models. The reported results show that training on all sources improves HF emulation and calibrated-LF emulation, while calibration inference is accurate in the low-dimensional analytic case for a source without model-form error, but becomes non-identifiable and initialization-dependent in the higher-dimensional engineering problem.

Significance. The strengths of the paper are its use of external ground truth in both examples, its handling of sources with different numbers of calibration parameters and multiple outputs, and its candid discussion of non-identifiability and initialization sensitivity. If the emulation gains are reproducible and robust, the architecture could be a useful contribution to multi-fidelity data fusion. However, the central attribution is not established: the reported gains from 'all sources' are not separated from having more training data or from the network's bias-correction capacity, and no comparison to any existing calibration method is provided. Because the paper's own conclusion restricts justification to 'maximizing emulation accuracy,' the title's question remains effectively open. The significance is therefore moderate and conditional on additional ablation and baseline experiments.

major comments (5)
  1. [Section 3.1 and Eq. (2)] The text in Section 3.1 states that insensitivity of HF emulation to dummy calibration values is achieved 'by adding a term to the loss function,' but Eq. (2) contains no such term, and Section 5 describes a Jacobian-based dummy-insensitivity term as a scheme that was tried and abandoned. The mechanism that makes Table 1's HF-emulation accuracy insensitive to the dummy θ is therefore unspecified; either a term must be added to Eq. (2) and described in Section 3.2, or the text must be revised to explain how masking alone enforces this insensitivity.
  2. [Section 4, Tables 1, 4, 5] The main evidence for the abstract's claim that 'our method can improve predictive accuracy' is the comparison of two-source versus all-source training in Tables 1, 4, and 5. This comparison conflates simultaneous calibration with two other factors: the availability of additional LF training data and the bias-correction capacity of Block 3. There is no ablation that trains the same multi-source network with calibration disabled (e.g., fixed θ and Lcal terms removed), nor a baseline that calibrates each LF source separately while keeping the shared multi-source emulator. Without such an ablation, the observed improvement cannot be attributed to the simultaneous-calibration mechanism that is the paper's subject.
  3. [Section 4.2, Table 3] The engineering calibration inference is unreliable: the estimated ν is 0.397 versus the true 0.260, the estimated σ0 for s2 is 767 MPa versus the true 650 MPa, the text reports that 'most of the Voce hardening parameters are estimated to be at the extrema of their sampling ranges,' and the estimates vary with initialization. Because the paper's motivating question is about calibration, and the high-dimensional experiment is the realistic test, the current results imply that the method's calibration inference fails exactly where it matters. The positive claims in the abstract and conclusion should be restricted to emulation accuracy, or the method must be modified and re-evaluated before it can be presented as a calibration framework.
  4. [Section 4, Tables 1–5] All quantitative results are from a single training run, yet Section 5 states that 'the model is very complex and its performance depends on initialization' and Section 4.2 reports that calibration estimates vary with initialization. Single-run RRMSE values in Tables 1, 4, and 5 cannot support comparative statements such as 'never the worst performing' or 'significantly improves' without seed sensitivity or repeated-run statistics. The authors should report means and standard deviations over multiple initializations, at least for the main tables.
  5. [Section 2] The paper does not compare iPro-NC against any existing method, including Kennedy-O'Hagan calibration, Latent Map GP [11], or the probabilistic neural data-fusion framework [37] on which the architecture relies. Even if the contribution is an exploratory framework, at least one baseline comparison is needed to interpret 'improve predictive accuracy' relative to prior work; without it, the reader cannot judge whether simultaneous calibration offers any advantage over calibrated separate emulation.
minor comments (6)
  1. [Table 4] In the first block of Table 4, the second row is labeled 's0 ands1' but appears intended to be 's0 ands2'; the duplicated label makes the table misleading.
  2. [Section 4.1, Table 1] The statement that 'the network is never the worst performing on any output' is a weak claim because in Table 1 all-source training is not best for y1 (0.0825 vs 0.0677 for s0 and s2); the phrasing should quantify differences or avoid suggestive causal language.
  3. [Eqs. (2)–(6)] The interval-score terms are written inconsistently: Eqs. (5) and (6) omit the output subscript i that is used in Eqs. (3) and (4), which makes the summation over outputs in Eq. (2) unclear.
  4. [Section 4.2] The footnote 'We fix εp in both the Holloman and Voce models' conflicts with Eqs. (13) and (14), where σy depends on εp as a variable; this needs rewording or clarification.
  5. [Figure 4] The captions for subfigures (a) and (b) are nearly identical ('Calibration LS for s2'); the reader would benefit from explicitly stating that (a) uses two sources and (b) uses three sources.
  6. [Section 3.2] The clause 'We consider the divergence only between the standard deviations' is vague; it should state that LKL in Eq. (9) penalizes only the estimated standard deviations and not the means.

Circularity Check

0 steps flagged · score 0.0 of 10

No material circularity: the central claims are tested against external analytic and FE ground truth, and the self-citations to prior PNDF work are design choices, not load-bearing derivations.

full rationale

The paper's central empirical claims do not reduce to their inputs by construction. The calibration loss (Eqs. 2-6) directly optimizes the match between HF data and LF predictions at the estimated parameters, so the later observation that posteriors cover theta_MSE (Section 4.1, Figure 2) is a consistency check on the optimization rather than an independent prediction; this is inherent to any calibration method and does not make the paper's contribution circular. The paper's headline claim ('Our method can improve predictive accuracy') is supported by out-of-sample RRMSE comparisons in Tables 1, 4, and 5 against known analytic source functions (Eqs. 10-12) and a fixed Holloman-law HF simulator, i.e., external benchmarks, not by the fitted values themselves. The latent-variable conversion is explicitly 'motivated by [37]' (Section 3.1), a same-author prior work, and the interval score is also taken from [37]; however, this is an architectural adoption rather than an appeal to a uniqueness theorem, and the paper does not cite it as a proof of its empirical results. The paper's own admissions of non-identifiability ('iPro-NC has very large learning capacity') are limitation statements, and the skeptic's concern about the missing no-calibration ablation is a causal-attribution gap (correctness risk), not an equation-level circularity. No step in the derivation, Eq. X = Eq. Y by definition, is exhibited.

Assumptions & free parameters 3 free parameters · 5 assumptions · 3 invented entities

The framework rests on a representational assumption (latent variable structure), on proper scoring rules, on the reparameterization trick, and on several ad hoc modeling choices (dummy calibration values for the HF source, two-step calibration split). The experiments use synthetic and FE data with known targets, but the method lacks theoretical guarantees and the paper reports no repeated-seed statistics.

free parameters (3)
  • beta_IS
    Weight of interval-score loss terms in Eq. 2; tuned by hand, value not reported.
  • beta_KL
    Weight of KL regularization term in Eq. 2; tuned by hand, value not reported.
  • sigma_p
    Prior standard deviation for calibration parameter distributions in Eq. 9; must be set from domain knowledge or tuned, value not reported.
assumptions (5)
  • domain assumption The relationship between all data sources can be represented by a latent variable model parameterized by neural networks.
    Section 3.1 states the network 'converts MF modeling and calibration to a latent variable modeling problem' based on [37].
  • standard math The interval score is a suitable proper scoring rule for training and evaluating prediction intervals.
    Section 3.2 uses LIS in the loss; proper scoring rules are standard for probabilistic predictions.
  • standard math The reparameterization trick yields unbiased gradients for the learned calibration parameter distributions.
    Section 3.3 relies on the reparameterization trick to train the network.
  • ad hoc to paper The HF source has no calibration parameters and can be emulated using dummy calibration values that the network must learn to ignore.
    Section 3.1 states dummy values are used for theta and a loss term enforces insensitivity; this is a modeling choice specific to iPro-NC.
  • ad hoc to paper In the engineering problem, calibration can be split into separate elastic and plastic steps, fixing some parameters in each step.
    Section 4.2 splits the calibration process into two steps based on material behavior; this improves performance but is tailored to the problem.
invented entities (3)
  • zs (source latent variable)
    purpose: Encodes similarities between data sources and is intended for visualizable source relationships.
    Figure 4c shows the learned source embedding is incorrect (s1 is placed closer to s0 than s2), so the interpretability claim is not externally validated.
  • ztheta (calibration parameter latent variable)
    purpose: Captures the effect of calibrated parameters on each LF source's output.
    Depends on non-identifiable calibration estimates; no independent handle is provided.
  • zc (categorical latent variable)
    purpose: Encodes relationships between categorical input combinations.
    Not directly tested in the experiments; serves only as an internal representation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Should We Simultaneously Calibrate Multiple Computer Models?." pith.science (2026). https://pith.science/paper/SYZ3672M

@misc{pith2026250518176,
  author       = {Pith},
  title        = {Pith review of: Should We Simultaneously Calibrate Multiple Computer Models?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SYZ3672M}},
  note         = {Machine review of arXiv:2505.18176}
}
read the original abstract

In an increasing number of applications designers have access to multiple computer models which typically have different levels of fidelity and cost. Traditionally, designers calibrate these models one at a time against some high-fidelity data (e.g., experiments). In this paper, we question this tradition and assess the potential of calibrating multiple computer models at the same time. To this end, we develop a probabilistic framework that is founded on customized neural networks (NNs) that are designed to calibrate an arbitrary number of computer models. In our approach, we (1) consider the fact that most computer models are multi-response and that the number and nature of calibration parameters may change across the models, and (2) learn a unique probability distribution for each calibration parameter of each computer model, (3) develop a loss function that enables our NN to emulate all data sources while calibrating the computer models, and (4) aim to learn a visualizable latent space where model-form errors can be identified. We test the performance of our approach on analytic and engineering problems to understand the potential advantages and pitfalls in simultaneous calibration of multiple computer models. Our method can improve predictive accuracy, however, it is prone to non-identifiability issues in higher-dimensional input spaces that are normally constrained by underlying physics.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 38 canonical work pages

  1. [11]

    Data fusion with latent map gaussian processes

    Jonathan Tammer Eweis-Labolle, Nicholas Oune, and Ramin Bostanabad. Data fusion with latent map gaussian processes. Journal of Mechanical Design, 144(9):091703, 2022

    work page 2022

  2. [37]

    Probabilistic neural data fusion for learning from an arbitrary number of multi-fidelity data sets

    Carlos Mora, Jonathan Tammer Eweis-Labolle, Tyler Johnson, Likith Gadde, and Ramin Bostanabad. Probabilistic neural data fusion for learning from an arbitrary number of multi-fidelity data sets. Com- puter Methods in Applied Mechanics and Engineering, 415:116207, 2023

    work page 2023

  3. [1]

    Welch, Toby J

    Jerome Sacks, William J. Welch, Toby J. Mitchell, and Henry P. Wynn. Design and Analysis of Computer Experiments. Statistical Science, 4(4):409–423, November 1989. Publisher: Institute of Mathematical Statistics

    work page 1989

  4. [2]

    Forest, Bruno Sans ´o, and Daniel Zantedeschi

    Chris E. Forest, Bruno Sans ´o, and Daniel Zantedeschi. Inferring climate system properties using a computer model. Bayesian Analysis, 3(1):1–37, March 2008. Publisher: International Society for Bayesian Analysis

    work page 2008

  5. [3]

    Salter, Daniel B

    James M. Salter, Daniel B. Williamson, John Scinocca, and Viatcheslav Kharin. Uncertainty Quan- tification for Computer Models With Spatial Output Using Calibration-Optimal Bases, October 2019. Publisher: Taylor & Francis

    work page 2019

  6. [4]

    Huseby, Bernard J

    Thorjørn Larssen, Ragnar B. Huseby, Bernard J. Cosby, Gudmund Høst, Tore Høg ˚asen, and Magne Aldrin. Forecasting Acidification Effects Using a Bayesian Calibration and Uncertainty Propagation Approach. Environmental Science & Technology, 40(24):7841–7847, December 2006

    work page 2006

  7. [5]

    Arhonditsis, Song S

    George B. Arhonditsis, Song S. Qian, Craig A. Stow, E. Conrad Lamon, and Kenneth H. Reckhow. Eutrophication risk assessment using Bayesian calibration of process-based models: Application to a mesotrophic lake. Ecological Modelling, 208(2-4):215–229, November 2007

    work page 2007

  8. [6]

    Henderson, Richard J

    Daniel A. Henderson, Richard J. Boys, Kim J. Krishnan, Conor Lawless, and Darren J. Wilkinson. Bayesian Emulation and Calibration of a Stochastic Computer Model of Mitochondrial DNA Deletions in Substantia Nigra Neurons. Journal of the American Statistical Association, 104(485):76–87, March 2009

    work page 2009

Show all 43 references
  1. [7]

    Combining experimental data and computer simulations, with an application to flyer plate experiments

    Jim Gattiker, Dave Higdon, Sallie Keller-McNulty, Michael McKay, Leslie Moore, and Brian Williams. Combining experimental data and computer simulations, with an application to flyer plate experiments. Bayesian Analysis, 1(4):765–792, December 2006. Publisher: International Soc...

    2006

  2. [8]

    Machine learning assisted calibration of a ductile fracture locus model

    Sandra Baltic, Mohammad Zhian Asadzadeh, Patrick Hammer, Julien Magnien, Hans-Peter G ¨anser, Thomas Antretter, and Ren´e Hammer. Machine learning assisted calibration of a ductile fracture locus model. Materials & Design, 203:109604, May 2021

    2021

  3. [9]

    ABAQUS/Standard User’s Manual, Version 6.9

    Michael Smith. ABAQUS/Standard User’s Manual, Version 6.9 . Dassault Syst `emes Simulia Corp, United States, 2009

    2009

  4. [10]

    Gaussian Processes For Machine Learning

    Carl Rasmussen and Christopher Williams. Gaussian Processes For Machine Learning . The MIT Press, 2006

    2006

  5. [12]

    Data-driven calibration of mul- tifidelity multiscale fracture models via latent map gaussian process

    Shiguang Deng, Carlos Mora, Diran Apelian, and Ramin Bostanabad. Data-driven calibration of mul- tifidelity multiscale fracture models via latent map gaussian process. Journal of Mechanical Design, 145(1):011705, 2023

    2023

  6. [13]

    GP+: A Python Library for Kernel-based learning via Gaussian Processes, December 2023

    Amin Yousefpour, Zahra Zanjani Foumani, Mehdi Shishehbor, Carlos Mora, and Ramin Bostan- abad. GP+: A Python Library for Kernel-based learning via Gaussian Processes, December 2023. arXiv:2312.07694 [cs, stat]. 19

    2023 arXiv

  7. [14]

    Extrapolation With Gaussian Random Pro- cesses and Evolutionary Programming, 2020

    Robert Planas, Nicholas Oune, and Ramin Bostanabad. Extrapolation With Gaussian Random Pro- cesses and Evolutionary Programming, 2020

    2020

  8. [15]

    Bayesian calibration of computer models

    Marc C Kennedy and Anthony O’Hagan. Bayesian calibration of computer models. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 63(3):425–464, 2001

    2001

  9. [16]

    Combining field data and computer simulations for calibration and prediction

    Dave Higdon, Marc Kennedy, James C Cavendish, John A Cafeo, and Robert D Ryne. Combining field data and computer simulations for calibration and prediction. SIAM Journal on Scientific Computing, 26(2):448–466, 2004

    2004

  10. [17]

    Bayesian calibration of inexact computer models

    Matthew Plumlee. Bayesian calibration of inexact computer models. Journal of the American Statis- tical Association, 112(519):1274–1285, 2017

    2017

  11. [18]

    Weizhao Zhang, Ramin Bostanabad, Biao Liang, Xuming Su, Danielle Zeng, Miguel A Bessa, Yan- chao Wang, Wei Chen, and Jian Cao. A numerical bayesian-calibrated characterization method for multiscale prepreg preforming simulations with tension-shear coupling.Composites Science a...

    2019

  12. [19]

    Understanding the effects of model uncertainty in robust design with computer experiments, 2006

    Daniel W Apley, Jun Liu, and Wei Chen. Understanding the effects of model uncertainty in robust design with computer experiments, 2006

    2006

  13. [20]

    A framework for validation of computer models

    Maria J Bayarri, James O Berger, Rui Paulo, Jerry Sacks, John A Cafeo, James Cavendish, Chin-Hsu Lin, and Jian Tu. A framework for validation of computer models. Technometrics, 49(2):138–154, 2007

    2007

  14. [21]

    Quantification of model uncertainty: Calibration, model discrepancy, and identifiability, 2012

    Paul D Arendt, Daniel W Apley, and Wei Chen. Quantification of model uncertainty: Calibration, model discrepancy, and identifiability, 2012

    2012

  15. [22]

    Improving identifiability in model calibration using multiple responses, 2012

    Paul D Arendt, Daniel W Apley, Wei Chen, David Lamb, and David Gorsich. Improving identifiability in model calibration using multiple responses, 2012

    2012

  16. [23]

    Uncertainty in predictions of the climate response to rising levels of greenhouse gases

    David A Stainforth, Tolu Aina, Carl Christensen, Mat Collins, Nick Faull, Dave J Frame, Jamie A Kettleborough, S Knight, A Martin, JM Murphy, et al. Uncertainty in predictions of the climate response to rising levels of greenhouse gases. Nature, 433(7024):403–406, 2005

    2005

  17. [24]

    Calibrating a large computer experiment simulating radiative shock hydrodynamics

    Robert B Gramacy, Derek Bingham, James Paul Holloway, Michael J Grosskopf, Carolyn C Kuranz, Erica Rutter, Matt Trantham, and R Paul Drake. Calibrating a large computer experiment simulating radiative shock hydrodynamics. The Annals of Applied Statistics, 9(3):1141–1168, 2015

    2015

  18. [25]

    Prediction based on the kennedy-o’hagan calibration model: asymptotic consis- tency and other properties, 2017

    Rui Tuo and CF Wu. Prediction based on the kennedy-o’hagan calibration model: asymptotic consis- tency and other properties, 2017

    2017

  19. [26]

    Adjustments to computer models via projected kernel calibration

    Rui Tuo. Adjustments to computer models via projected kernel calibration. SIAM/ASA Journal on Uncertainty Quantification, 7(2):553–578, 2019

    2019

  20. [27]

    Improving identifiability in model calibration using multiple responses

    Paul D Arendt, Daniel W Apley, Wei Chen, David Lamb, and David Gorsich. Improving identifiability in model calibration using multiple responses. Journal of Mechanical Design, 134(10):100909, 2012

    2012

  21. [28]

    Preposterior analysis to select experimental responses for improving identifiability in model uncertainty quantification

    Zhen Jiang, Wei Chen, and Daniel W Apley. Preposterior analysis to select experimental responses for improving identifiability in model uncertainty quantification. In ASME 2013 International Design Engineering Technical Conferences and Computers and Information in Engineering ...

    2013

  22. [29]

    Computer model calibration using high dimensional output, 2008

    Dave Higdon, James Gattiker, Brian Williams, and Maria Rightley. Computer model calibration using high dimensional output, 2008

    2008

  23. [30]

    Dropout as a bayesian approximation: Representing model uncer- tainty in deep learning

    Yarin Gal and Zoubin Ghahramani. Dropout as a bayesian approximation: Representing model uncer- tainty in deep learning. In Proceedings of the 33rd International Conference on International Confer- ence on Machine Learning - Volume 48, ICML’16, page 1050–1059. JMLR.org, 2016

    2016

  24. [31]

    Srivastava, G

    N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov. Dropout: A simple way to prevent neural networks from overfitting.Journal of Machine Learning Research, 15(1):1929–1958, 2014

    1929

  25. [32]

    Variational physics-informed neu- ral networks for solving partial differential equations, 2019

    Ehsan Kharazmi, Zhongqiang Zhang, and George Em Karniadakis. Variational physics-informed neu- ral networks for solving partial differential equations, 2019

    2019

  26. [33]

    Weight uncertainty in neural network

    Charles Blundell, Julien Cornebise, Koray Kavukcuoglu, and Daan Wierstra. Weight uncertainty in neural network. In International conference on machine learning, pages 1613–1622. PMLR, 2015

    2015

  27. [34]

    Deep learning

    Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. Deep learning. Nature, 521(7553):436–444, May

  28. [35]

    Hamel, Kevin N

    Craig M. Hamel, Kevin N. Long, and Sharlotte L. B. Kramer. Calibrating constitutive models with full-field data via physics informed neural networks. Strain, 59(2):e12431, April 2023. Publisher: John Wiley & Sons, Ltd

    2023

  29. [36]

    Simultaneous and meshfree topology optimization with physics-informed gaussian processes, 2024

    Amin Yousefpour, Shirin Hosseinmardi, Carlos Mora, and Ramin Bostanabad. Simultaneous and meshfree topology optimization with physics-informed gaussian processes, 2024

    2024

  30. [38]

    Deep evidential regres- sion

    Alexander Amini, Wilko Schwarting, Ava Soleimany, and Daniela Rus. Deep evidential regres- sion. In Proceedings of the 34th International Conference on Neural Information Processing Systems, NIPS’20, pages 14927 – 14937, Red Hook, NY , USA, 2020. Curran Associates Inc

    2020

  31. [39]

    Decoupled Weight Decay Regularization, January 2019

    Ilya Loshchilov and Frank Hutter. Decoupled Weight Decay Regularization, January 2019. arXiv:1711.05101 [cs, math]

    2019 arXiv

  32. [40]

    Auto-encoding variational bayes, 2013

    Diederik P Kingma and Max Welling. Auto-encoding variational bayes, 2013

    2013

  33. [41]

    Pytorch lightning, March 2019

    William Falcon and The PyTorch Lightning team. Pytorch lightning, March 2019. license: ”Apache- 2.0”, repository-code: ”https://github.com/Lightning-AI/lightning”, version: ”1.4”

    2019

  34. [42]

    Determination of hardness-strength and -flow behavior relationships in bulged aluminum alloys and verification by FE analysis on Rockwell hardness test

    Amir Asgharzadeh, Sobhan Alah Nazari Tiji, Rasoul Esmaeilpour, Taejoon Park, and Farhang Pour- boghrat. Determination of hardness-strength and -flow behavior relationships in bulged aluminum alloys and verification by FE analysis on Rockwell hardness test. The International Jo...

    2020

  35. [2015]

    Publisher: Nature Publishing Group

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.