pith. sign in

arxiv: 2605.04051 · v1 · submitted 2026-02-28 · 📊 stat.ML · cs.LG· math.OC

A Consistency-Centric Approach to Set-Based Optimization with Multiple Models of Unranked Fidelity

Danielle F. Morey , Giulia Pedrielli , Cherry Y. Wakayama , Zelda B. Zabinsky This is my paper

Pith reviewed 2026-05-15 17:50 UTC · model grok-4.3

classification 📊 stat.ML cs.LGmath.OC
keywords set-based optimizationmultiple modelsmodel consistencyfidelity uncertaintyprobabilistic boundsmulti-fidelity optimizationS-BOMM
0
0 comments X

The pith

A set-based method finds reliable optimization solutions by measuring consistency across multiple models without ranking their fidelity.

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

The paper introduces S-BOMM, a methodology for optimization problems involving several models of unknown relative accuracy. It shifts focus from identifying the single best solution under one presumed high-fidelity model to finding solutions that perform consistently well across all available models. A probabilistic analysis bounds the chance that the consistency-based selection yields correct or incorrect outcomes. Tests on benchmark problems show that this approach locates good solutions even when no model is designated as superior.

Core claim

S-BOMM identifies sets of solutions that exhibit high consistency in their performance across multiple models of unranked fidelity. By prioritizing agreement between models rather than alignment with any single one, the method produces candidate solutions whose quality is supported by cross-model agreement, with theoretical bounds on error rates derived from the consistency measure.

What carries the argument

The consistency metric that quantifies agreement in solution quality across models, used to select sets rather than single points.

If this is right

  • Solutions selected for high cross-model consistency are more likely to remain effective when the true system differs from every individual model.
  • The provided probability bounds let users set reliability targets by choosing how many models to include or how many evaluations to run.
  • Optimization can proceed in domains where no single model can be validated or labeled highest-fidelity in advance.
  • Empirical behavior on test problems shows improved recovery of good solutions relative to single-model baselines when multiple plausible models exist.

Where Pith is reading between the lines

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

  • The same consistency principle could apply to ensemble modeling that combines outputs from independent sources or different approximation levels.
  • In engineering practice the approach may reduce reliance on costly high-fidelity runs by accepting results where cheaper models already agree.
  • Adaptive evaluation strategies could add new model runs only in regions where current consistency is low, refining the selected set with minimal extra cost.
  • Tighter bounds might be obtained by incorporating known correlations or shared structure among the models.

Load-bearing premise

Consistency between models of unranked fidelity reliably indicates solution quality without external validation against ground truth or any model ranking.

What would settle it

On a problem with a known ground-truth function plus several inaccurate models, if S-BOMM repeatedly selects solutions that are poor under ground truth yet highly consistent across the inaccurate models, the central claim is falsified.

Figures

Figures reproduced from arXiv: 2605.04051 by Cherry Y. Wakayama, Danielle F. Morey, Giulia Pedrielli, Zelda B. Zabinsky.

Figure 1
Figure 1. Figure 1: S-BOMM methodology. the bounded decision space. Let there be N different models, fn(x) with x ∈ X and n = 1, . . . , N. When branching, let σi be the i-th subregion in X, where the subregions have no intersection and cover the entirety of the decision space, that is, σi ∩ σj = ∅ for all i ̸= j and S i σi = X. Given K classes, a user-selected classifier assigns one of the K possible classes to sub￾region σi… view at source ↗
Figure 2
Figure 2. Figure 2: Impact of v and r on the probability of consistently classifying correctly as in (5), consistently classifying incorrectly as in (6), and inconsistently classifying as in (7) 14 [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Heatmap of the objective functions of the three models (9), (10), (11) in Example 1 [PITH_FULL_IMAGE:figures/full_fig_p016_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Example 1 target regions (a) and solution set (b). [PITH_FULL_IMAGE:figures/full_fig_p016_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Heatmap of the objective function of the four models (12), (13), (14), (15) in [PITH_FULL_IMAGE:figures/full_fig_p019_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Example 2 target regions (a) and solution set (b). [PITH_FULL_IMAGE:figures/full_fig_p019_6.png] view at source ↗
read the original abstract

In complex real-world settings, optimization is challenged by the presence of diverse models of differing fidelity. In many optimization problems, a single model is treated as the most accurate representation of the underlying system, while other models are evaluated primarily by their agreement with this presumed most accurate model. Yet in real-world applications, model accuracy is rarely known a priori and assuming a single most accurate model can be misleading. This paper addresses this gap by proposing a flexible set-based optimization methodology called Set-Based Optimization with Multiple Models (S-BOMM) that works with multiple models without the assumption of a most accurate high-fidelity model. Unlike traditional optimization approaches that focus on finding an optimal solution according to the high-fidelity model, our methodology utilizes consistency between models to identify good solutions across multiple models. A probabilistic analysis of the consistency method is provided that bounds the likelihood of the methodology producing correct or incorrect results. Empirical results demonstrate the effectiveness of S-BOMM on test problems. By focusing on the consistency across models rather than relying on a single best solution, this set-based approach offers a practical alternative to optimization problems where multiple models must be considered without assuming a single most accurate high-fidelity model.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript proposes Set-Based Optimization with Multiple Models (S-BOMM), a methodology that identifies good solutions via consistency across multiple models of unranked fidelity rather than designating a single most-accurate high-fidelity model. It supplies a probabilistic analysis that bounds the likelihood of correct versus incorrect outcomes and reports empirical effectiveness on test problems.

Significance. If the probabilistic bounds are shown to hold when model errors may be correlated, the work would provide a practical alternative for optimization under model uncertainty, particularly in engineering and scientific domains where fidelity rankings are unavailable a priori. The consistency-centric framing avoids a common but often unrealistic modeling assumption.

major comments (1)
  1. [Abstract] Abstract (probabilistic analysis paragraph): the claimed bounds on the probability of correct/incorrect results are not shown to address correlated model errors. If models share a common bias, high consistency will reinforce the shared error; the analysis must either derive the bound under an independence assumption that is stated explicitly or provide a robustness argument that survives correlation. This assumption is load-bearing for the central claim that consistency reliably indicates solution quality.
minor comments (2)
  1. The abstract states empirical effectiveness on test problems but supplies no details on problem dimensions, number of models, or how ground-truth correctness was defined for the reported bounds; these omissions hinder immediate assessment of the experiments.
  2. Notation for the consistency metric and the set-construction rule should be introduced with a small illustrative example before the general formulation to improve readability.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their thoughtful review and for identifying a key assumption in our probabilistic analysis. We address the comment below and have revised the manuscript to make the independence assumption explicit while noting its implications.

read point-by-point responses

  1. Referee: [Abstract] Abstract (probabilistic analysis paragraph): the claimed bounds on the probability of correct/incorrect results are not shown to address correlated model errors. If models share a common bias, high consistency will reinforce the shared error; the analysis must either derive the bound under an independence assumption that is stated explicitly or provide a robustness argument that survives correlation. This assumption is load-bearing for the central claim that consistency reliably indicates solution quality.

    Authors: We agree that the analysis requires an explicit statement of its assumptions. The probabilistic bounds are derived under the assumption that model errors are independent; this is used to bound the probability that multiple models agree on an incorrect solution. We have revised the abstract to state this assumption directly and added a paragraph in Section 3.2 of the manuscript discussing the role of independence. We also note that correlated errors (e.g., shared bias) would invalidate the current bounds and could lead to over in incorrect solutions; a full robustness analysis under arbitrary correlations is left for future work. The revised text makes the load-bearing assumption transparent without claiming robustness beyond independence. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper proposes S-BOMM as a set-based method that identifies solutions via cross-model consistency and separately supplies a probabilistic analysis bounding the chance of correct versus incorrect outcomes. The consistency metric and the bounding analysis are distinct; the bounds are presented as derived from model-error assumptions rather than by re-expressing the input consistency measure as the output. No equation or claim reduces by construction to its own inputs, no self-citation is load-bearing for the central result, and the derivation remains self-contained against external benchmarks. This is the normal, non-circular outcome for a methodological proposal that separates its heuristic from its performance guarantee.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The approach rests on the domain assumption that cross-model consistency can substitute for fidelity ranking; no free parameters or invented physical entities are described in the abstract.

axioms (2)
  • domain assumption Consistency between models of unranked fidelity identifies good solutions
    Central premise stated in the abstract as the basis for identifying solutions.
  • domain assumption Probabilistic bounds on correctness can be derived from the consistency method
    Invoked to support the methodology's reliability claims.

pith-pipeline@v0.9.0 · 5526 in / 1287 out tokens · 59070 ms · 2026-05-15T17:50:20.161865+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

45 extracted references · 45 canonical work pages

  1. [1]

    Aarnes, J. E., V. L. Hauge, and Y. Efendiev (2007). Coarsening of three-dimensional struc- tured and unstructured grids for subsurface flow.Advances in Water Resources 30(11), 2177–2193

    work page 2007

  2. [2]

    Alam, K. M. and A. E. Saddik (2017). C2PS: A digital twin architecture reference model for the cloud-based cyber-physical systems.IEEE Access 5, 2050–2062.https://doi. org/10.1109/ACCESS.2017.2657006

    work page doi:10.1109/access.2017.2657006 2017

  3. [3]

    Banks, J. and J. S. Carson (1986). Introduction to discrete-event simulation. In J. Wilson, J. Henriksen, and S. Roberts (Eds.),Proceedings of the Winter Simulation Conference 1986, WSC ’86, New York, NY, USA, pp. 17–23. Association for Computing Machinery. https://doi.org/10.1145/318242.318253

    work page doi:10.1145/318242.318253 1986

  4. [4]

    Barton, R. R. (1994). Metamodeling: A state of the art review. In J. D. Tew, S. Mani- vannan, D. A. Sadowski, and A. F. Seila (Eds.),Proceedings of the Winter Simulation Conference 1994, Lake Buena Vista, FL, pp. 237–244. Institute of Electrical and Elec- tronics Engineers Inc.https://doi.org/10.1109/WSC.1994.717134

    work page doi:10.1109/wsc.1994.717134 1994

  5. [5]

    Barton, R. R. and M. Meckesheimer (2006). Metamodel-based simulation optimization. In S. G. Henderson and B. L. Nelson (Eds.),Simulation, Volume 13 ofHandbooks in Operations Research and Management Science, pp. 535–574. Elsevier.https://doi. org/10.1016/S0927-0507(06)13018-2

    work page doi:10.1016/s0927-0507(06)13018-2 2006

  6. [6]

    Bramerdorfer, G., J. A. Tapia, J. J. Pyrhonen, and A. Cavagnino (2018). Modern electrical machine design optimization: Techniques, trends, and best practices.IEEE Transac- tions on Industrial Electronics 65, 7672–7684.https://doi.org/10.1109/TIE.2018. 2801805. 22

    work page doi:10.1109/tie.2018 2018

  7. [7]

    Cao, H., F. Li, G. Zeng, X. Wang, C. Dong, and M. Zhang (2025). Quadratic polynomial regression model based on l2 regularization. In2025 IEEE International Conference on

    work page 2025

  8. [8]

    Choi, S. H., K. M. Seo, and T. G. Kim (2017). Accelerated simulation of discrete event dy- namic systems via a multi-fidelity modeling framework.Applied Sciences (Switzerland) 7, 1–23.https://doi.org/10.3390/app7101056

    work page doi:10.3390/app7101056 2017

  9. [9]

    Conn, A. R., K. Scheinberg, and L. N. Vicente (2009).Introduction to Derivative-Free

    work page 2009

  10. [10]

    Costabal, F. S., P. Perdikaris, E. Kuhl, and D. E. Hurtado (2019). Multi-fidelity clas- sification using Gaussian processes: Accelerating the prediction of large-scale compu- tational models.Computer Methods in Applied Mechanics and Engineering 357, 1–16. https://doi.org/10.1016/j.cma.2019.112602

    work page doi:10.1016/j.cma.2019.112602 2019

  11. [11]

    Filipowicz, B. and J. Kwiecie´ n (2008). Queueing systems and networks. models and appli- cations.Bulletin of the Polish Academy of Sciences: Technical Sciences 56(4), 379– 390.https://www.researchgate.net/publication/267403147_Queueing_systems_ and_networks_Models_and_applications

    work page arXiv 2008

  12. [12]

    Gnanasambandam, R., B. Shen, A. C. C. Law, C. Dou, and Z. J. Kong (2025). Deep Gaus- sian process for enhanced Bayesian optimization and its application in additive manu- facturing.IISE Transactions 57(4), 423–436.https://doi.org/10.1080/24725854. 2024.2312905

    work page doi:10.1080/24725854 2025

  13. [13]

    Jaenisch, H. M. and J. W. Handley (2011). A robust regularization algorithm for polynomial networks for machine learning. In M. Blowers, T. H. O’Donnell, and O. L. Mendoza- Schrock (Eds.),Evolutionary and Bio-Inspired Computation: Theory and Applications V, Volume 8059, pp. 1–21. International Society for Optics and Photonics: SPIE.https: //doi.org/10.111...

    work page doi:10.1117/12.884284 2011

  14. [14]

    Lee, S., Z. B. Zabinsky, J. N. Wasserheit, S. M. Kofsky, and S. Liu (2021). COVID-19 pandemic response simulation in a large city: Impact of nonpharmaceutical interventions on reopening society.Medical Decision Making 41, 419–429.https://doi.org/10. 1177/0272989X2110030

    work page 2021

  15. [15]

    Li, K. and F. Li (2024). Multi-fidelity methods for optimization: A survey. arXiv preprint arXiv:2402.09638 [cs.LG],https://doi.org/10.48550/arXiv.2402.09638

    work page doi:10.48550/arxiv.2402.09638 2024

  16. [16]

    Li, P. and S. Chen (2018). Hierarchical Gaussian processes model for multi-task learning. Pattern Recognition 74, 134–144.https://doi.org/10.1016/j.patcog.2017.09.021

    work page doi:10.1016/j.patcog.2017.09.021 2018

  17. [17]

    Linz, D. D., H. Haung, and Z. B. Zabinsky (2017). Multi-fidelity simulation optimiza- tion with level set approximation using probabilistic branch and bound. In W. K. V

    work page 2017

  18. [18]

    D’Ambrogio, G

    Chan, A. D’Ambrogio, G. Zacharewicz, N. Mustafee, G. Wainer, and E. Page (Eds.), Proceedings of the Winter Simulation Conference 2017, Las Vegas, NV, pp. 2057–2068. Institute of Electrical and Electronics Engineers Inc.https://doi.org/10.1109/WSC. 2017.8247939. 23

    work page doi:10.1109/wsc 2017

  19. [19]

    Liao, and J

    Liu, J., H. Liao, and J. A. White (2021). Queueing analysis of the replenishment of multiple in-the-aisle pick positions.IISE Transactions 53(1), 1–20.https://doi.org/10.1080/ 24725854.2020.1731773

    work page arXiv 2021

  20. [20]

    Macal, C. M. and M. J. North (2009). Agent-based modeling and simulation. In M. D

    work page 2009

  21. [21]

    Rossetti, R. R. Hill, B. Johansson, A. Dunkin, and R. G (Eds.),Proceedings of the Winter Simulation Conference 2009, Austin, TX, pp. 86–98. Institute of Electrical and Electronics Engineers Inc.https://doi.org/10.1109/WSC.2009.5429318

    work page doi:10.1109/wsc.2009.5429318 2009

  22. [22]

    MacKay, D. J. C. (1998). Introduction to Gaussian processes.NATO ASI series F computer and systems sciences 168, 133–166

    work page 1998

  23. [23]

    Morey, D. F., R. S. Plate, C. Y. Wakayama, and Z. B. Zabinsky (2024). Multifidelity topol- ogy design of a maritime survey operation with UUVs.Frontiers in Marine Science 11, 1–14.https://doi.org/10.3389/fmars.2024.1277719

    work page doi:10.3389/fmars.2024.1277719 2024

  24. [24]

    Morey, D. F., Z. B. Zabinsky, C. Wakayama, and R. Plate (2021). Multi-fidelity modeling for the design of a maritime environmental survey network utilizing unmanned underwater vehicles. In S. Kim, B. Feng, K. Smith, S. Masoud, Z. Zheng, C.Szabo, and M. Loper (Eds.),Proceedings of the Winter Simulation Conference 2021, Volume 2021-December,

    work page 2021

  25. [25]

    Phoenix, AX, pp. 1–12. Institute of Electrical and Electronics Engineers Inc.https: //doi.org/10.1109/WSC52266.2021.9715365

    work page doi:10.1109/wsc52266.2021.9715365 2021

  26. [26]

    Opitz, D. and R. Maclin (1999). Popular ensemble methods: An empirical study.Journal of Artificial Intelligence Research 11, 169–198.https://doi.org/10.1613/jair.614

    work page doi:10.1613/jair.614 1999

  27. [27]

    Katariya, and V

    Parmar, A., R. Katariya, and V. Patel (2018). A review on random forest: An en- semble classifier. InProceedings of the International Conference on Intelligent Data Communication Technologies and Internet of Things 2018, Volume 26, pp. 758–763. https://doi.org/10.1007/978-3-030-03146-6_86

    work page doi:10.1007/978-3-030-03146-6_86 2018

  28. [28]

    Khandait, Y

    Pedrielli, G., T. Khandait, Y. Cao, Q. Thibeault, H. Huang, M. Castillo-Effen, and G. Fainekos (2024). Part-X: A family of stochastic algorithms for search-based test gen- eration with probabilistic guarantees.IEEE Transactions on Automation Science and Engineering 21(3), 4504–4525. 10.1109/TASE.2023.3297984

    work page doi:10.1109/tase.2023.3297984 2024

  29. [29]

    Peherstorfer, K

    Peherstorfer, B., K. Willcox, and M. Gunzburger (2018). Survey of multifidelity methods in uncertainty propagation, inference, and optimization.SIAM Review 60, 550–591. https://doi.org/10.1137/16M1082469

    work page doi:10.1137/16m1082469 2018

  30. [30]

    Sanchez, S. M. and T. W. Lucas (2002). Exploring the world of agent-based simulations: Simple models. complex analyses. In E. Yucesan, C.-H. Chen, J. L. Snowdon, and J. M. Charnes (Eds.),Proceedings of the Winter Simulation Conference 2002, San Diego, CA, pp. 116–126. Institute of Electrical and Electronics Engineers Inc.https://doi.org/ 10.1109/WSC.2002.1172875

    work page doi:10.1109/wsc.2002.1172875 2002

  31. [31]

    Hawkins-Daarud, A

    Sel, K., A. Hawkins-Daarud, A. Chaudhuri, D. Osman, A. Bahai, D. Paydarfar, K. Will- cox, C. Chung, and R. Jafari (2025). Survey and perspective on verification, valida- tion, and uncertainty quantification of digital twins for precision medicine.npj Digital Medicine 8(40), 1–12.https://doi.org/10.1038/s41746-025-01447-y. 24

    work page doi:10.1038/s41746-025-01447-y 2025

  32. [32]

    Simpson, T. W., J. D. Peplinski, P. N. Koch, and J. K. Allen (2001). Metamodels for computer-based engineering design: Survey and recommendations.Engineering with Computers 17, 129–150

    work page 2001

  33. [33]

    Tan, B. and A. Matta (2024). The digital twin synchronization problem: Framework, formulations, and analysis.IISE Transactions 56(6), 652–665.https://doi.org/10. 1080/24725854.2023.2253869

    work page arXiv 2024

  34. [34]

    Tolles, J. and T. Luong (2020). Modeling epidemics with compartmental models.JAMA Guide to Statistics and Methods 323(24), 2515–2516.https://doi.org/doi:10.1001/ jama.2020.8420. van Zyl, J. P. and A. P. Engelbrecht (2023). Set-based particle swarm optimisation: A review.Mathematics 11, 1–36.https://doi.org/10.3390/math11132980

    work page doi:10.3390/math11132980 2020

  35. [35]

    Walter, G. G. and M. Contreras (1999).Compartmental modeling with networks. Boston, MA: Springer Science & Business Media

    work page 1999

  36. [36]

    Wan, X. (2019). The effect of regularization coefficient on polynomial regression.Journal of Physics: Conference Series 1213(4), 1–4.https://doi.org/10.1088/1742-6596/ 1213/4/042054

    work page doi:10.1088/1742-6596/ 2019

  37. [37]

    Xia, Q., T. Shi, M. Y. Wang, and S. Liu (2010). A level set based method for the optimization of cast part.Structural and Multidisciplinary Optimization 41, 735–747. https://doi.org/10.1007/s00158-009-0444-7

    work page doi:10.1007/s00158-009-0444-7 2010

  38. [38]

    Zhang, E

    Xu, J., S. Zhang, E. Huang, C. H. Chen, L. H. Lee, and N. Celik (2015). Efficient multi- fidelity simulation optimization. In A. Tolk, S. Y. Diallo, I. O. Ryzhov, L. Yilmaz, S. Buckley, and J. A. Miller (Eds.),Proceedings of the Winter Simulation Conference 2015, Volume 2015-January, Savannah, GA, pp. 3940–3951. Institute of Electrical and Electronics Eng...

    work page doi:10.1109/wsc.2014.7020219 2015

  39. [39]

    Zhang, E

    Xu, J., S. Zhang, E. Huang, C. H. Chen, L. H. Lee, and N. Celik (2016). MO2TOS: Multi-fidelity optimization with ordinal transformation and optimal sampling.Asia- Pacific Journal of Operational Research 33, 1–26.https://doi.org/10.1142/ S0217595916500172

    work page 2016

  40. [40]

    Schmidt and L

    Yin, X., S. Bushaj, Y. Yuan, and ˙I. E. B¨ uy¨ uktahtakın (2024). COVID-19: Agent-based simulation-optimization to vaccine center location vaccine allocation problem.IISE Transactions 56(7), 699–714.https://doi.org/10.1080/24725854.2023.2223246

    work page doi:10.1080/24725854.2023.2223246 2024

  41. [41]

    Zabinsky, Z. B. and H. Huang (2020). A partition-based optimization approach for level set approximation: Probabilistic branch and bound. InWomen in Industrial and Systems

    work page 2020

  42. [42]

    Zabinsky, Z. B., G. Pedrielli, and H. Huang (2019). A framework for multi-fidelity modeling in global optimization approaches. In P. N. G.Pardalos, R. Umeton, G. Giuffrida, and V. Sciacca (Eds.),from Proceedings of the Machine Learning, Optimization, and Data Science 5th International Conference, LOD 2019, Volume 11943 LNCS, pp. 335–346. Springer.https://...

    work page doi:10.1007/978-3-030-37599-7_28 2019

  43. [43]

    Zabinsky, Z. B., M. E. Tuttle, and C. Khompatraporn (2006).A Case Study: Composite Structure Design Optimization, Volume 85, pp. 507–528. Boston, MA: Springer US. https://doi.org/10.1007/0-387-30927-6_21

    work page doi:10.1007/0-387-30927-6_21 2006

  44. [44]

    2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition , volume =

    Zhang, Y. and D.-Y. Yeung (2010). Multi-task warped Gaussian process for personalized age estimation. InProceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2010, San Francisco, CA, pp. 2622–2629. Institute of Electrical and Electronics Engineers Inc.https://doi.org/10.1109/CVPR.2010.5539975

    work page doi:10.1109/cvpr.2010.5539975 2010

  45. [45]

    Zhang, Z., Z. Guan, Y. Gong, D. Luo, and L. Yue (2022). Improved multi-fidelity simulation-based optimisation: application in a digital twin shop floor.International Journal of Production Research 60, 1016–1035.https://doi.org/10.1080/00207543. 2020.1849846. 26

    work page doi:10.1080/00207543 2022