Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Posterior sampling with Adaptive Gaussian Processes in Bayesian parameter identification

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that interleaving posterior sampling with joint optimization of evaluation positions and tolerances substantially reduces the cost of surrogate-based Bayesian posterior sampling.

desk verdict Careful extension of adaptive GP surrogates with a real KL error bound, but the headline savings claim is only demonstrated under simulated Gaussian solver noise, not real FE solves. read the letter →

arxiv 2411.17858 v1 pith:Q5IIPZVO submitted 2024-11-26 math.NA cs.NA

Paolo Villani , Daniel Andrés-Arcones , Jörg F. Unger , Martin Weiser This is my paper · ORCID
classification math.NAcs.NA MSC 62F1560G1565C6065N21
keywords BayesianinverseproblemsGaussianprocessregressionadaptivedesignofexperimentsposteriorsamplingsurrogatemodelsevaluationtoleranceMarkovchainMonteCarlouncertaintyquantification
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 argues that, when a Gaussian-process surrogate stands in for an expensive forward model in a Bayesian inverse problem, the surrogate should be built and the posterior sampled as one interleaved process, not as separate phases. Its central claim is that greedily choosing new evaluation points by their sensitivity to posterior error and simultaneously re-optimizing how accurately old and new points are evaluated, all under a fixed work budget, substantially reduces the computational effort needed to sample the posterior compared with adapting positions only or using a static space-filling design. The efficiency gain is demonstrated on three forward models (synthetic, diffusion, and Poisson) in 2, 3, and 4 parameter dimensions, with discretization error emulated by Gaussian noise. If the claim holds, practitioners with expensive PDE solvers could obtain faithful uncertainty estimates with fewer solver runs by spending accuracy only where the posterior places mass.

What carries the argument

The load-bearing object is the surrogate-based posterior $\pi(p\mid y^m,D)$ of Eq. (7), in which the Gaussian-process predictive distribution $y_D(p)\sim N(\bar y(p),\Gamma(p))$ enters the likelihood through Eq. (6); including the predictive variance prevents the surrogate from overconfidently narrowing the posterior. Around this object the paper builds two computable error models, the GP-averaged KL upper bound (13) and the posterior-weighted $L^2$ error (14), and a work model $W(\tau)=\tau^{-l/r}$ for adaptive finite-element solves. The greedy mechanism is Algorithm 1: draw or update MCMC samples from the current surrogate posterior, use those samples to evaluate the acquisition function (20) for candidate positions, solve the tolerance optimization problem (21) under the remaining budget, and re-evaluate the forward model only at changed designs. The sliding-window recycling of samples is what makes the optimization problems cheap enough to solve at every iteration.

What would settle it

Run the same algorithm on a genuine adaptive finite-element problem with a corner singularity, where discretization error is controlled by mesh refinement and is not independent Gaussian noise, and compare posterior error at the same budget; if the tolerance-optimizing variant does not beat position-only adaptation, the central efficiency claim fails.

Watch

Extended reading notes

Core claim

The paper's central discovery is a working formulation of this interleaved idea: it converts the unknown error between the true posterior and the surrogate-based posterior into computable error models, namely an upper bound on the Kullback-Leibler divergence (Theorem 1, Eq. (13)) and a posterior-weighted $L^2$ error model (Eq. (14)), and then uses these to drive a sequential greedy design. At each iteration, MCMC samples from the current surrogate-based posterior weight the acquisition function (20) that scores candidate evaluation points, and the same samples discretize the tolerance optimization problem (21) that reallocates the remaining budget among old and new points. The paper shows numerically that this joint adaptation beats position-only adaptation, that tolerance optimization is most valuable when more accurate solves are cheap relative to the work model $W(\tau)=\tau^{-l/r}$, and that the fully adaptive designs end with far fewer training points, so both forward evaluations and surrogate evaluations become cheaper.

Load-bearing premise

Everything hinges on the assumption that a forward-model evaluation with tolerance $\tau$ returns the true value plus independent zero-mean Gaussian noise with standard deviation $\tau$; real discretization errors from adaptive solvers are not independent or Gaussian, and the numerical experiments substitute Gaussian noise for actual solver error.

Editorial extensions

If this is right

  • Under the paper's error model, tolerance optimization plus position adaptation converges to the target posterior with fewer forward-model evaluations than position-only or static designs on the same budget.
  • Tolerance optimization pays off most when the work model exponent $l/r$ is small, i.e. when increasing evaluation accuracy is comparatively cheap; for expensive accurate solves, position adaptation nearly suffices.
  • Fully adaptive designs end with substantially fewer training points, making the surrogate itself cheaper to evaluate during sampling.
  • Using the stochastic likelihood (6) instead of plugging in the GP mean guards against overconfident posteriors in early iterations.
  • Reusing samples in a sliding window reduces sampling cost when the posterior changes little between iterations.

Reading between the lines

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

  • Because the experiments emulate discretization error as independent zero-mean Gaussian noise with variance $\tau^2$, the reported efficiency gains are best read as an upper bound for real adaptive PDE solvers, whose discretization errors are correlated and not Gaussian; the acquisition rule may misallocate budget in that regime.
  • Tolerance re-optimization assumes evaluations can be stored and continued from previous grids; for black-box solvers that cannot resume a solve, the method loses much of its advantage even if the error model held.
  • A natural testable extension is to treat the reported tolerance $\tau_i$ as a tunable hyperparameter or to model correlated error components, and to compare the greedy acquisition (20) against a fully Bayesian design under the same budget.
  • The bound (13) sets $\alpha=1$ optimistically once the surrogate is reasonable, so in early iterations the true bound may be larger and practical convergence could depend on the initial design more than the asymptotic analysis suggests.
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

4 major / 4 minor

Summary. This paper proposes a fully adaptive Gaussian-process surrogate strategy for sampling Bayesian posteriors. Starting from an initial design, Algorithm 1 interleaves MCMC sampling from the current surrogate-based posterior with greedy design updates: candidate evaluation points are selected as local maximizers of the acquisition function (20), and evaluation tolerances are re-optimized in (21) under a work-model constraint. Two error models are derived, the KL-divergence model (13) based on Theorem 1 and Lemma 1, and the posterior-weighted L2 model (14). Numerical experiments on a 2D synthetic example, a 3D diffusion example, and a 4D Poisson example compare AGP-const and AGP-geom with position-only-adaptive (posAGP) and static LHSGP designs, reporting improved convergence and fewer training points. The abstract's central claim is that joint optimization of positions and tolerances significantly reduces computational effort.

Significance. Strengths: the fully stochastic likelihood (6) is a principled treatment of surrogate uncertainty; Theorem 1 and Lemma 1 provide explicit, analytically derived error bounds; Appendix A gives closed-form derivatives, making the acquisition and tolerance optimization concrete and reproducible; the experimental matrix across dimensions, cost models, and error metrics is extensive. If the observed gains survive contact with genuine adaptively-refined PDE solvers, the method would be a practical improvement over position-only-adaptive surrogates. However, the current numerical evidence is produced under the Gaussian-noise idealization of Section 3.1, not under the FE discretization errors the method is motivated by, so the significance for the intended application is not yet established.

major comments (4)
  1. [5.1; Remark 1] The central numerical claim in the abstract, that jointly optimizing evaluation positions and tolerances yields a significant reduction of computational effort, is supported only under the idealized error model of Section 3.1, in which evaluation errors are independent N(0, tau_i^2) draws. Section 5.1 implements this by adding Gaussian noise to analytic forward models; no actual adaptive FE solver is run. Because FE discretization errors are deterministic, generally biased, and spatially correlated (as Remark 1 concedes), the tolerance-optimization steps in (18) and (21) use the wrong likelihood for the intended application. A concrete test would be to run Algorithm 1 with an actual adaptive FE solve, or with a deterministic error model having known, correlated error, and to report whether the observed reduction over posAGP and LHSGP persists; without such a test the abstract's claim is overstated.
  2. [4.1.1, Eq. (13)] Theorem 1 assumes pi_D(y_m) <= alpha pi(y_m) for some finite alpha, but the error model sets log alpha = 0 'optimistically'. Since L_D(p) is the convolution of the measurement error with the GP predictive density, the evidence ratio pi_D(y_m)/pi(y_m) is not generally bounded by 1, especially when the surrogate mean is biased. Dropping log alpha systematically underestimates e^KL_D(p) and can change the selected positions and tolerances in (20) and (21). The authors should either justify alpha = 1 from the setting, estimate the ratio, or report the sensitivity of the resulting designs to the value of log alpha.
  3. [5.2-5.4, Eq. (15)] The work model W(tau) = tau^{-l/r} is assumed and not measured, and the reported savings are expressed in units of this assumed model rather than in wall-clock time or actual solver effort. Table 1 and Table 2 report numbers of training points, but the adaptive strategies also spend computational effort on multi-start optimization for (20), SLSQP for (21), and MCMC sampling between iterations; this overhead is not included in the comparison. To substantiate the claimed reduction in computational effort, the paper should report the overhead of the design optimization and, ideally, total runtime for each strategy.
  4. [4.3.1, Eq. (18)] The acquisition function and the tolerance optimization in (19) replace the posterior pi(.|y_m,D_j) by the currently available pi(.|y_m,D_{j-1}) and freeze the predictive mean at the previous iterate. These substitutions are acknowledged, but they are not covered by Theorem 1, so the resulting error model is an ad-hoc heuristic whose failure modes are not assessed. The numerical section should include a sensitivity check, for example comparing the designs produced by the heuristic against a small number of random design perturbations, or against an oracle that evaluates (8) with the exact posterior, to show that the design decisions are robust to this approximation.
minor comments (4)
  1. [3.1] The sentence 'Gaussian process Regression (GPR) is an approximation technique that can efficiently learn large classes of functions [23], providing a .' contains an incomplete final phrase; the missing object should be supplied.
  2. [Algorithm 1, step 3] The notation r_j is used before it is defined, and the final design is written as D_j = ((p_i,tau_i))_{i=1,...,s_j+1}, which does not clearly account for candidates excluded by tau_i = infinity; please clarify the indexing.
  3. [Figures 5, 6, 10] The convergence plots do not label their axes; please state explicitly whether the horizontal axis is iteration number or accumulated work and what the vertical metric is (e.g., E^KL or E^{L2}).
  4. [References] Reference [30] is cited only as 'In Proceedings Of The Conference Algoritmy, 2024'; please provide complete bibliographic information.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the efficiency claim is a new empirical comparison conditional on an explicitly stated idealized error model; prior self-citations are contextual, not load-bearing.

full rationale

The derivation chain is self-contained rather than circular. The error models (13) and (14) are GP-averaged upper bounds derived in Theorem 1, Lemma 1 and Corollaries 1-2; although they are evaluated with the current surrogate posterior and predictive variance Gamma(p), they serve only as acquisition and tolerance-selection heuristics in Algorithm 1 (Eqs. (20)-(21)), while the reported convergence rates are measured as actual errors against the analytically available forward models (Sec. 5.1), so the central claim is not identical to the optimized surrogate objective. The method extends the authors' earlier work [30], and the acquisition rule cites [30, 26], but the algorithm and derivative formulas are fully specified here rather than importing an unverified premise, and no uniqueness theorem or forbidden alternative is invoked. The work model W(tau)=tau^(-l/r) (Eq. (15)) is an assigned asymptotic model, not a fitted parameter, so no quantity is fitted and then renamed as a prediction. The paper itself flags in Remark 1 that independent Gaussian evaluation errors are 'for sure not satisfied' for FE discretization errors, and Sec. 5.1 simulates discretization error by adding Gaussian noise to analytic models; this is an explicitly acknowledged limitation affecting transferability of the efficiency claim, but it does not make any derived result equivalent to its inputs by construction.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the idealized evaluation-noise model (Gaussian errors with known variance), the asymptotic work model W(tau)=tau^(-l/r), and the optimistic choice alpha=1 in the error bound. The Gaussian process prior and Bayesian measurement model are standard.

free parameters (2)
  • Geometric budget ratio alpha (AGP-geom) = 1.173 (2D), 1.178 (3D), 1.148 (4D)
    Chosen a priori to define the geometric budget schedule; it affects the pace at which computational budget is spent but is not fitted to the results.
  • Initial tolerance tau for default evaluations = 0.05 (2D), 0.02 (3D), 0.04 (4D)
    Hand-chosen for the initial design; the non-adaptive benchmarks use these fixed tolerances, so the comparison budget depends on them.
assumptions (5)
  • domain assumption Forward model evaluations are corrupted by independent, zero-mean Gaussian noise with standard deviation equal to the tolerance tau_i.
    Used in the GP regression likelihood (Section 3.1); Remark 1 acknowledges it is not satisfied for actual finite element discretization errors.
  • domain assumption The computational cost of an evaluation depends only on the tolerance via W(tau)=tau^(-l/r).
    Section 4.2 cites [25,32]; it is asymptotic and ignores position dependence (Remark 3).
  • ad hoc to paper The ratio pi_D(y_m)/pi(y_m) is bounded by alpha, and alpha may be set to 1.
    Theorem 1 requires a finite bound alpha, but Section 4.1.1 sets alpha=1 optimistically, which may make the derived upper bound invalid; the error model is then used as a heuristic.
  • standard math The forward model y is a realization of a Gaussian process GP(mu,k) with known kernel.
    This is the standard GP regression assumption (Section 3.1); it is not an unconventional postulate.
  • domain assumption Evaluation results can be stored and refined, and grids can be reused across parameter points.
    Remark 4 in Section 4.2 assumes that saving the final solution and grid allows efficient tolerance refinement, which is natural but not guaranteed for all PDE solvers.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Posterior sampling with Adaptive Gaussian Processes in Bayesian parameter identification." pith.science (2026). https://pith.science/paper/Q5IIPZVO

@misc{pith2026241117858,
  author       = {Pith},
  title        = {Pith review of: Posterior sampling with Adaptive Gaussian Processes in Bayesian parameter identification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q5IIPZVO}},
  note         = {Machine review of arXiv:2411.17858}
}
read the original abstract

Posterior sampling by Monte Carlo methods provides a more comprehensive solution approach to inverse problems than computing point estimates such as the maximum posterior using optimization methods, at the expense of usually requiring many more evaluations of the forward model. Replacing computationally expensive forward models by fast surrogate models is an attractive option. However, computing the simulated training data for building a sufficiently accurate surrogate model can be computationally expensive in itself, leading to the design of computer experiments problem of finding evaluation points and accuracies such that the highest accuracy is obtained given a fixed computational budget. Here, we consider a fully adaptive greedy approach to this problem. Using Gaussian process regression as surrogate, samples are drawn from the available posterior approximation while designs are incrementally defined by solving a sequence of optimization problems for evaluation accuracy and positions. The selection of training designs is tailored towards representing the posterior to be sampled as good as possible, while the interleaved sampling steps discard old inaccurate samples in favor of new, more accurate ones. Numerical results show a significant reduction of the computational effort compared to just position-adaptive and static designs.

Figures

Figures reproduced from arXiv: 2411.17858 by the authors.

Figure 1
Figure 1. Surrogate mean error and training points (black) for different steps of a realization of [PITH_FULL_IMAGE:figures/full_fig_p012_1.png] view at source ↗
Figure 2
Figure 2. Ground truth forward model for the 2-d parameter space. [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗
Figure 3
Figure 3. Accumulated samples (black line), drawn samples (green bar), discarded samples (red bar) [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Distribution of the logarithm of the evaluation tolerances for [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: Convergence rates for the two different error models and four different FE simulation costs [PITH_FULL_IMAGE:figures/full_fig_p015_5.png]
Figure 6
Figure 6. Figure 6: Convergence rates for the two different error models and two different FE simulation costs [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]
Figure 7
Figure 7. Figure 7: Comparison of the samples from the different posteriors in one run of the 3-d experiment. [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]
Figure 8
Figure 8. Figure 8: Considered solution and sensors. For the initial design D0, 17 initial evalu￾ation points are selected by Latin Hypercube Sampling with default tolerance τ = 0.04. Ev￾ery iteration we seek up to 5 local maximizers of the acquisition function as candidates and we perfor…
Figure 9
Figure 9. Figure 9: Accumulated samples (black line), drawn samples (green bar), discarded samples (red bar) [PITH_FULL_IMAGE:figures/full_fig_p019_9.png]
Figure 10
Figure 10. Figure 10: Convergence rates for the two different error models and two different FE simulation [PITH_FULL_IMAGE:figures/full_fig_p020_10.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Robust Surrogate-Based Bayesian Inference via Sampling-Based Adaptive Active Learning (SALE)

    stat.CO 2026-08 conditional novelty 8.0 of 10

    Using the expected posterior as a common design measure for both optimization and uncertainty reduction yields robust finite-budget posterior approximations.

Reference graph

Works this paper leans on

36 extracted references · 35 canonical work pages · cited by 1 Pith paper

  1. [1]

    T. Bai, A. L. Teckentrup, and K. C. Zygalakis. Gaussian processes for bayesian inverse problems associated with linear partial differential equations. Statistics and Computing , 34, 2024

    work page 2024

  2. [2]

    G. Boehl. An ensemble mcmc sampler for robust bayesian inference. Technical report, University of Bonn, 2024

    work page 2024

  3. [3]

    J. A. Christen and C. Fox. Markov chain monte carlo using an approximation. Journal of Computational and Graphical Statistics , 14(4):795–810, 2005

    work page 2005

  4. [4]

    Christmann and I

    A. Christmann and I. Steinwart. Support Vector Machines. Springer New York, NY, 2008

    work page 2008

  5. [5]

    Crombecq, E

    K. Crombecq, E. Laermans, and T. Dhaene. Efficient space-filling and non-collapsing sequential design strategies for simulation-based modeling. European Journal of Operational Research , 214:683–696, 2011

    work page 2011

  6. [6]

    Dinkel, C

    M. Dinkel, C. M. Geitner, G. R. Rei, J. Nitzler, and W. A. Wall. Solving bayesian inverse problems with expensive likelihoods using constrained gaussian processes and active learning. Inverse Problems, 40(9):095008, jul 2024

    work page 2024

  7. [7]

    D. K. Duvenaud. Automatic Model Construction with Gaussian Processes . Phd thesis, Univer- sity of Cambridge, 2014

    work page 2014

  8. [8]

    H.W. Engl, M. Hanke, and A. Neubauer. Regularization of inverse problems . Kluwer, 1996. 20

    work page 1996

Show all 36 references
  1. [9]

    Foreman-Mackey

    D. Foreman-Mackey. corner.py: Scatterplot matrices in python. The Journal of Open Source Software, 1(2):24, jun 2016

    2016

  2. [10]

    Foreman-Mackey, D

    D. Foreman-Mackey, D. W. Hogg, D. Lang, and J. Goodman. emcee: The mcmc hammer. PASP, 125:306–312, 2013

    2013

  3. [11]

    Gpy- torch: Blackbox matrix-matrix gaussian process inference with gpu acceleration

    Jacob Gardner, Geoff Pleiss, Kilian Q Weinberger, David Bindel, and Andrew G Wilson. Gpy- torch: Blackbox matrix-matrix gaussian process inference with gpu acceleration. In S. Bengio, H. Wallach, H. Larochelle, K. Grauman, N. Cesa-Bianchi, and R. Garnett, editors, Advances in ...

    2018

  4. [12]

    Giunta, S

    A. Giunta, S. Wojtkiewicz, and M. Eldred. Overview of modern design of experiments methods for computational simulations. In 41st Aerospace Sciences Meeting and Exhibit, AIAA 2003- 649, pages 1–17, 2003

    2003

  5. [13]

    Goodman and J

    J. Goodman and J. Weare. Ensemble samplers with affine invariance. Communications in applied mathematics and computational science , 5(1):65–80, 2010

    2010

  6. [14]

    Hastings

    W.K. Hastings. Monte Carlo sampling methods using Markov chains and their applications. Biometrika, 57(1):97–109, 1970

    1970

  7. [15]

    Helin, A.M

    T. Helin, A.M. Stuart, A.L. Teckentrup, and K.C. Zygalakis. Introduction to gaussian process regression in bayesian inverse problems, with new results on experimental design for weighted error measures. In Monte Carlo and Quasi-Monte Carlo Methods , 2023

    2023

  8. [16]

    Hennig and C.J

    P. Hennig and C.J. Schuler. Entropy search for information-efficient global optimization.Journal of Machine Learning Research, 13:1809–1837, 2012

    2012

  9. [17]

    Joseph and Y

    V. Joseph and Y. Hung. Orthogonal-maximin latin hypercube designs. Statistica Sinica, 18:171– 186, 2008

    2008

  10. [18]

    J¨ arvenp¨ a¨ a, M

    M. J¨ arvenp¨ a¨ a, M. U. Gutmann, A. Vehtari, and P. Marttine. Parallel Gaussian process surrogate Bayesian inference with noisy likelihood evaluations. Bayesian Analysis, 16, pp. 147–178. , 2021

    2021

  11. [19]

    Kaipio and E

    J. Kaipio and E. Somersalo. Statistical and Computational Inverse Problems . Springer, 2005

    2005

  12. [20]

    Lehmensiek, P

    R. Lehmensiek, P. Meyer, and M. M¨ uller. Adaptive sampling applied to multivariate, multi- ple output rational interpolation models with application to microwave circuits. International Journal of RF and Microwave Computer-Aided Engineering , 12(4):332–340, 2002

    2002

  13. [21]

    Moˇ ckus

    J. Moˇ ckus. On Bayesian methods for seeking the extremum. In Optimization Techniques IFIP Technical Conference Novosibirsk, pages 400–404. Springer, 1975

    1975

  14. [22]

    Queipo, R

    N. Queipo, R. Haftka, W. Shyy, T. Goel, R. Vaidyanathan, and P. Tucker. Surrogate-based analysis and optimization. Progress in Aerospace Sciences, 41(1):1–28, 2005

    2005

  15. [23]

    Rasmussen and C.K.I

    C. Rasmussen and C.K.I. Williams. Gaussian Processes for Machine Learning . MIT Press, 2006

    2006

  16. [24]

    Schneider, M

    P. Schneider, M. Hammerschmidt, L. Zschiedrich, and S. Burger. Using Gaussian process regression for efficient parameter reconstruction. In Metrology, Inspection, and Process Control for Microlithography XXXIII, volume 10959. SPIE, 2019

    2019

  17. [25]

    Semler and M

    P. Semler and M. Weiser. Adaptive Gaussian process regression for efficient building of surrogate models in inverse problems. Inverse Problems, 39:125003, 2023

    2023

  18. [26]

    Semler and M

    P. Semler and M. Weiser. Adaptive gradient enhanced gaussian process surrogates for inverse problems. In Proceedings of the MATH+ Thematic Einstein Semester 2023 , 2024

    2023

  19. [27]

    Sinsbeck and W

    M. Sinsbeck and W. Nowak. Sequential design of computer experiments for the solution of bayesian inverse problems. SIAM/ASA Journal on Uncertainty Quantification , 5(1):640–664, 2017. 21

    2017

  20. [28]

    Sugiyama

    M. Sugiyama. Active learning in approximately linear regression based on conditional expecta- tion of generalization error. Journal of Machine Learning Research , 7:141––166, 2006

    2006

  21. [29]

    Ter Braak

    C.J.F. Ter Braak. A markov chain monte carlo version of the genetic algorithm differential evolu- tion: easy bayesian computing for real parameter spaces. Statistics and Computing, 16:239–249, 2006

    2006

  22. [30]

    Villani, J

    P. Villani, J. F. Unger, and M. Weiser. Adaptive gaussian process regression for bayesian inverse problems. In Proceedings Of The Conference Algoritmy, 2024

    2024

  23. [31]

    Wang and M

    Z. Wang and M. Broccardo. A novel active learning-based gaussian process metamodelling strategy for estimating the full probability distribution in forward uq analysis. Struct. Safety , 84:101937, 2020

    2020

  24. [32]

    Weiser and S

    M. Weiser and S. Ghosh. Theoretically optimal inexact spectral deferred correction methods. Commu. Appl. Math. Comp. Sci. , 13(1):53–86, 2018

    2018

  25. [33]

    J. Wu, M. Poloczek, A.G. Wilson, and P.I. Frazier. Bayesian optimization with gradients. In Advances in Neural Information Processing Systems , volume 30, pages 3–6. Curran Associates, 2017

    2017

  26. [34]

    A. Zaytsev. Reliable surrogate modeling of engineering data with more than two levels of fidelity. In 2016 7th International Conference on Mechanical and Aerospace Engineering (ICMAE) , pages 341–345, 2016

    2016

  27. [35]

    H. Zhu, R. H. Byrd, and J. Nocedal. Algorithm 778: L-bfgs-b, fortran routines for large scale bound constrained optimization. ACM Transactions on Mathematical Software , 23:550–560, 1997

    1997

  28. [36]

    ´Alvarez, L

    M.A. ´Alvarez, L. Rosasco, and N.D. Lawrence. Kernels for vector-valued functions: A review. Foundations and Trends in Machine Learning , 4(3):195–266, 2012. A Target function derivatives This appendix contains the analytical expression of the derivatives contained in the acqu...

    2012

Pith tools

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