Pith. sign in

REVIEW 3 major objections 4 minor 32 references

Feasibility-Driven Trust Region Bayesian Optimization

T0 review · 3 major / 4 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 FuRBO, a feasibility-driven trust-region Bayesian optimization method, matches or beats state-of-the-art constrained optimizers on benchmark suites and finds feasible solutions where competitors fail, especially when feasible regions are…

desk verdict FuRBO is a solid incremental improvement over SCBO for rare-feasibility constrained BO; the inspector-based trust region is the real contribution, and the experiments support it despite some presentational overreach. read the letter →

arxiv 2506.14619 v1 pith:HX6XGDVV submitted 2025-06-17 cs.LG

Paolo Ascia , Elena Raponi , Thomas Bäck , Fabian Duddeck This is my paper · ORCID
classification cs.LG
keywords Bayesianoptimizationconstrainedblack-boxtrustregioninspectorsamplingGaussianprocesssurrogateThompsonBBOB-constrainedbenchmarkrarefeasibility
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

The paper introduces FuRBO, a Bayesian optimization method for black-box problems with expensive, unknown constraints, aimed at settings where merely finding a feasible point is hard. Rather than centering the trust region only on the best evaluated point, as the SCBO algorithm does, FuRBO samples a set of 'inspector' points uniformly in a ball around the current best, ranks them with Gaussian-process surrogate models of the objective and constraints, and builds its trust region as the smallest hyperrectangle containing the top-ranked inspectors. Thompson sampling then proposes new candidates inside that region, and the region's radius doubles or halves with progress, letting the search refocus rapidly. The paper reports that FuRBO ties SCBO on easy problems and outperforms it as constraints become numerous and hard to satisfy, and on physics-inspired benchmarks it is comparable or better, with its clearest wins where feasible regions are rare.

What carries the argument

The carrying mechanism is the inspector-defined trust region: at each iteration, a population of inspector points is sampled uniformly inside a ball of radius $R$ centered at the current best candidate, ranked by surrogate predictions of objective value and constraint violation, and the top $P\%$ of inspectors are enclosed by the smallest axis-aligned hyperrectangle, which becomes the trust region for Thompson sampling. The ranking metric orders feasible samples by objective value and infeasible samples by maximum normalized constraint violation, so the trust region points toward the part of the landscape the surrogates judge most likely to contain high-quality feasible solutions. The radius $R$ is doubled on success and halved on failure, letting the search refocus aggressively.

What would settle it

On a synthetic problem whose feasible region is a narrow curved corridor lying entirely outside the initial ball of radius $R$ around the current best, run FuRBO and a global random-search baseline for the same budget; if FuRBO returns no feasible point while random search does, the claim that inspector-guided trust regions accelerate feasibility discovery is refuted.

Watch

Extended reading notes

Core claim

The central claim is that the trust region itself can be made feasibility-driven: instead of using the best feasible point only as a center, FuRBO samples inspector points uniformly in a ball of radius $R$ around the current best, ranks them using the GP surrogate models of objective and constraints, and defines the next trust region as the axis-aligned hyperrectangle containing the top $P\%$ of inspectors. Thompson sampling then selects new candidates within this region. The paper argues that this procedure actively reuses constraint-model information to hunt for feasible territory, making the trust region shift and resize substantially between iterations. Empirically, the authors show that FuRBO either ties or outperforms SCBO, constrained EI, COBYLA, CMA-ES, and random search across the BBOB-constrained benchmark suite and physics-inspired problems, with the largest margins when constraints are numerous, active, or hard to satisfy, and in high dimensions where SCBO stops finding feasible solutions.

Load-bearing premise

The load-bearing premise is that the top-ranked inspector points, sampled in a ball around the current best and scored by Gaussian-process surrogates, outline a hyperrectangle that actually contains a promising feasible region.

Editorial extensions

If this is right

  • On the full BBOB-constrained suite in 10D, FuRBO converges faster and reaches lower final regret than SCBO on problems with 17 or more constraints, while performing comparably on problems with one to three constraints.
  • In 40D, FuRBO finds feasible solutions under a moderate constraint load ($9+\lfloor 3D/4\rfloor$ constraints) on nearly all benchmark functions, whereas SCBO systematically fails within the same budget.
  • On the 60D rover trajectory problem with 15 hard constraints, FuRBO reaches a higher reward than SCBO, and on the 30D Keane bump it slightly outperforms SCBO.
  • On low-dimensional, lightly constrained engineering problems such as spring, welded beam, and pressure vessel design, FuRBO matches SCBO without a notable advantage, indicating its benefit is concentrated in hard-feasibility settings.
  • Larger inspector percentages and very small initial radii degrade FuRBO's performance, so its gains depend on the choice of these hyperparameters.

Reading between the lines

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

  • The paper does not explore this, but the inspector-ranking idea could be extended to a covariance-shaped trust region fit to the top inspectors, which would likely help when the feasible set is a narrow rotated corridor, a limitation the paper itself acknowledges.
  • A testable prediction of the mechanism is that FuRBO's advantage over SCBO grows with the fraction of the budget spent before the first feasible point is found; a benchmark that varies only that fraction would isolate the trust-region effect.
  • Because the extra cost comes from surrogate evaluations of inspectors, FuRBO is most attractive when each true evaluation is expensive; on cheap problems the overhead may erase the gain, a tradeoff the paper notes only in passing.
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

3 major / 4 minor

Summary. The paper introduces FuRBO, a trust-region Bayesian optimization algorithm for constrained black-box problems. FuRBO extends SCBO by constructing the trust region from a set of 'inspector' points sampled uniformly in a ball around the current best solution, ranking them with surrogate models of the objective and constraints, and taking the smallest hyperrectangle containing the top P% of inspectors. The algorithm is evaluated on the BBOB-constrained COCO benchmark (3 instances in 2, 10, and 40 dimensions), on three representative functions against cEI, COBYLA, CMA-ES, and random search, and on additional benchmarks (Keane bump, spring, welded beam, pressure vessel, speed reducer, rover trajectory) against SCBO. The central claim is that FuRBO ties or outperforms state-of-the-art alternatives, with the largest gains when feasibility is rare and hard to locate.

Significance. If the empirical claims hold, FuRBO provides a practical improvement for constrained BO in settings where feasible regions are narrow. The paper's strengths include a publicly available implementation, statistical significance testing (Wilcoxon rank-sum) across replications, a systematic ablation study of four hyperparameters, and an honest limitations section that acknowledges failure in extremely constrained high-dimensional problems. The trend in Tables 1-3—growing advantage over SCBO as dimension and constraint count increase—is consistent with the proposed mechanism. However, the overstated evaluation scope ('full suite' vs 3 instances) and the internal budget inconsistency need correction before the claims can be taken at face value.

major comments (3)
  1. [Section 3 and Section 5.1] The problem definition in Section 3 states a total evaluation budget of 10D function evaluations, while every experimental setup in Section 5.1 uses a total evaluation budget of 30D (e.g., the Baselines Setup paragraph and the constrained BBOB paragraph). This internal inconsistency affects the framing of the method as a 'low-budget' optimizer and must be reconciled by either changing the problem definition or explicitly justifying the experimental budget.
  2. [Abstract and Section 5.1] The abstract and Section 5.1 claim that FuRBO is tested on 'the full BBOB-constrained COCO benchmark suite,' but the experimental protocol uses 3 instances (and 10 repetitions) rather than the 15 instances that constitute the full suite, in dimensions 2, 10, and 40 only. This overstates the evaluation scope; the authors should either run the full 15-instance suite or revise the wording to accurately describe the subset used.
  3. [Appendix E.3] The ablation study on the inspector percentage P in Appendix E.3 states that FuRBO's performance deteriorates as P increases, with the best results at P=1%, yet the main experiments fix P=10% without a stated reason. Given that the paper claims superiority partly through this feasibility-driven mechanism, the sensitivity of the algorithm to P should be addressed, either by justifying the chosen value across a wider set of problems or by tuning P per problem. As written, the choice appears to favor the reported results.
minor comments (4)
  1. [Section 5.2.1] The sentence 'For simpler problems (leftmost columns with 1–3 constraints)' is inaccurate because the leftmost columns in Table 2 correspond to 1, 3, and 9 constraints, not 1–3.
  2. [Section 5.2.1] The phrase 'configurations with 17 or more constraints' does not match the actual constraint counts in the rightmost columns (e.g., 16, 24, 54 constraints in 10D); please correct the threshold to align with the table.
  3. [Appendix A, Table 4 caption] The caption contains a typo: 'fro constrained optimization' should read 'for constrained optimization'.
  4. [Algorithm 1 and Section 4.2] The ranking metric r used in Algorithm 1 (lines 2 and 5) is only described in prose in Section 4.2; providing a formal numbered definition (including the normalization of constraint violations and the concatenation of feasible and infeasible samples) would significantly improve clarity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FuRBO is assessed on external benchmarks and its trust-region mechanism is an algorithmic heuristic, not a derivation that encodes the target results.

full rationale

FuRBO's central claim is empirical: inspector-guided trust regions improve constrained Bayesian optimization relative to SCBO and other baselines. The trust-region construction (Algorithm 1, lines 4-7) is a heuristic search mechanism; no optimality or feasibility theorem is derived from it, and the evaluation is against BBOB-constrained COCO functions, the Keane bump function, and physics-inspired benchmarks whose feasible sets and optima are defined externally and independently of FuRBO's outputs. The ranking metric r is used both to identify the current best point and to rank newly sampled inspectors, but this is algorithmic consistency rather than definitional circularity: the inspectors are fresh points, and the trust region is the smallest hyperrectangle over their top P%, which is not equal to the set of already-evaluated points selected by r. The hyperparameters (P=10%, R=1, tau_s=2, tau_f=3) were selected through ablations on one bent_cigar instance; this is a tuning and generalization concern, not a circularity, because the reported comparisons cover many functions, constraint levels, and dimensions beyond that instance, and no target result is encoded in the algorithm. The only self-citation (Raponi et al. 2019) is a background reference for crashworthiness applications and is not load-bearing. The paper's own limitation paragraphs (Section 6) acknowledge failures in extreme 40D settings, which scopes the claim honestly rather than circularly validating it. No equation in the paper reduces to its own inputs by construction, and no fitted parameter is renamed as a prediction.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on empirical benchmarking rather than derivation. Free parameters are hyperparameters set by hand; none are fitted to the benchmark outcomes in a way that encodes the result. The main structural assumption is that local inspector sampling and GP-ranked TRs can locate rare feasible regions.

free parameters (5)
  • Inspector percentage P = 10% (default)
    Fraction of top-ranked inspectors used to define the TR; ablation on bent_cigar shows performance deteriorates as P increases (Appendix E.3).
  • Initial sampling radius R = 1.0
    Radius of the uniform ball for inspectors; ablations show values below 0.5 degrade performance (Appendix E.2).
  • Success/failure thresholds tau_s, tau_f = 2 and 3
    Control how often the inspector sampling radius doubles or halves; fixed for all experiments (Section 5.1).
  • Total evaluation budget = 30D (main experiments)
    Problem definition states 10D budget but experiments use 30D; affects comparison to 'tight budget' claims (Section 3 vs 5.1).
  • Batch size q = 3D (main), 1 (baseline comparison)
    Batch size affects convergence; ablation in Appendix E.4.
assumptions (4)
  • domain assumption Gaussian process surrogates adequately approximate objective and constraint functions in the tested high-dimensional problems.
    Ranking of inspectors and Thompson sampling both rely on GP posterior quality (Section 4.2).
  • domain assumption The feasible region is reachable by inspectors sampled uniformly in a ball around the current best point.
    Inspectors are the only mechanism for moving the TR; if feasibility is distant, the TR never finds it (Algorithm 1, line 4).
  • domain assumption The BBOB-constrained suite and physics-inspired benchmarks are representative of real constrained black-box problems.
    Generalization claims are extrapolated from these benchmarks (Section 5).
  • standard math Standard probability theory and GP posterior sampling are valid.
    Used implicitly in GPs and Thompson sampling (Section 4.2, line 8).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Feasibility-Driven Trust Region Bayesian Optimization." pith.science (2026). https://pith.science/paper/HX6XGDVV

@misc{pith2026250614619,
  author       = {Pith},
  title        = {Pith review of: Feasibility-Driven Trust Region Bayesian Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HX6XGDVV}},
  note         = {Machine review of arXiv:2506.14619}
}
read the original abstract

Bayesian optimization is a powerful tool for solving real-world optimization tasks under tight evaluation budgets, making it well-suited for applications involving costly simulations or experiments. However, many of these tasks are also characterized by the presence of expensive constraints whose analytical formulation is unknown and often defined in high-dimensional spaces where feasible regions are small, irregular, and difficult to identify. In such cases, a substantial portion of the optimization budget may be spent just trying to locate the first feasible solution, limiting the effectiveness of existing methods. In this work, we present a Feasibility-Driven Trust Region Bayesian Optimization (FuRBO) algorithm. FuRBO iteratively defines a trust region from which the next candidate solution is selected, using information from both the objective and constraint surrogate models. Our adaptive strategy allows the trust region to shift and resize significantly between iterations, enabling the optimizer to rapidly refocus its search and consistently accelerate the discovery of feasible and good-quality solutions. We empirically demonstrate the effectiveness of FuRBO through extensive testing on the full BBOB-constrained COCO benchmark suite and other physics-inspired benchmarks, comparing it against state-of-the-art baselines for constrained black-box optimization across varying levels of constraint severity and problem dimensionalities ranging from 2 to 60.

Figures

Figures reproduced from arXiv: 2506.14619 by the authors.

Figure 1
Figure 1. One iteration of FuRBO. The leftmost panel shows the true objective and constraint isocon￾tours, with the global optimum in red. The next two panels show surrogate models of the objective (top) and aggregated constraint (bottom), built from evaluated points (black dots); the current best solution is marked in red. Inspectors (white crosses) are sampled around this point and ranked by feasibility and objective value.… view at source ↗
Figure 2
Figure 2. Loss convergence curve on the full constrained BBOB suite at 10D. Results are averaged across 3 instances with 10 repetitions each. The plot shows the mean loss with shaded areas indicating one standard error. FuRBO consistently outperforms SCBO on more severely constrained problems and performs comparably on easier ones. run with a batch size 𝑞 = 3𝐷, and total evaluation budget of 30𝐷 to mimic real-world scenarios … view at source ↗
Figure 3
Figure 3. Convergence comparison of FuRBO against SCBO, CEI, COBYLA, CMA-ES, and random sampling on 𝑓sphere, 𝑓bent_cigar, and 𝑓rast_rot in 10D. Curves show the mean loss over 10 repeti￾tions of the same instance, with shaded regions indicating one standard error. cantly better performance on the majority of the 10D problems. Similar figures to [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (15 more)
Figure 4
Figure 4. Figure 4: Loss convergence curve on 𝑓sphere, 𝑓bent_cigar, and 𝑓rast_rot constrained BBOB functions at 10D and mid-severe constraint level, for varying dimensionality of the problems. Results are averaged across 3 instances with 10 repetitions each. The plot shows the mean loss w…
Figure 5
Figure 5. Figure 5: Loss convergence curve on the full constrained BBOB suite at 2D. Results are averaged across 3 instances with 10 repetitions each. The plot shows the mean loss with shaded areas indicating one standard error. FuRBO and SCBO have comparable convergence trends. 17 [PITH…
Figure 6
Figure 6. Figure 6: Loss convergence curve on the full constrained BBOB suite at 40D. Results are averaged across 3 instances with 10 repetitions each. The plot shows the mean loss with shaded areas indicating one standard error. FuRBO is on par or outperforms SCBO for mild and medium-sev…
Figure 7
Figure 7. Figure 7: Average CPU time (in seconds) of FuRBO and SCBO across 10D all the constrained BBOB functions under increasing numbers of constraints. The two methods have comparable runtime, which increases with constraint severity. 19 [PITH_FULL_IMAGE:figures/full_fig_p019_7.png]
Figure 8
Figure 8. Figure 8: compares FuRBO and SCBO with other established baselines (CEI, COBYLA, and CMA￾ES) on three representative functions in 10D and medium-high constraint severity (9 + ⌊3𝐷/4⌋ constraints, 6 + ⌊𝐷/2⌋ active). As expected, the surrogate-based methods (CEI, SCBO, and FuRBO) a…
Figure 9
Figure 9. Figure 9: presents a study on the impact of the size of the initial sample set, also referred to as Design of Experiments (DoE), on the optimization performance of FuRBO. We compare four DoE sizes proportional to the problem dimensionality, specifically, 1D, 3D, 5D, and 10D init…
Figure 10
Figure 10. Figure 10: Study on the impact of the initial radius of the uniform distribution for the inspectors. The following initial radii are compared on the 𝑓bent_cigar BBOB function with 24 constraints: 𝑅 = {1.0, 0.5, 0.2, 0.1, 0.05}. The results show that the algorithm’s performance d…
Figure 11
Figure 11. Figure 11: Study on the impact of the percentage 𝑃% of inspectors selected to define the position and extension of the TR. The following percentages are compared: 1%, 5%, 10%, and 20%. Experiments run on 𝑓bent_cigar BBOB function with 24 constraints. 22 [PITH_FULL_IMAGE:figures…
Figure 12
Figure 12. Figure 12: Study on the impact of the batch size 𝑞 on the performance of FuRBO. The following configurations are compared: 1 (sequential), 1D, 2D, 3D, 4D, and 5D samples per batch. Experiments run on 𝑓bent_cigar BBOB function with 24 constraints. E.4 Batch size [PITH_FULL_IMAGE…
Figure 13
Figure 13. Figure 13: Convergence of FuRBO and SCBO on the 30D Keane Bump function under two constraints in the domain [0, 10] 𝐷 . Results are averaged over 10 independent runs for each algorithm. The plot shows the mean objective value, with shaded areas representing one standard error. F…
Figure 14
Figure 14. Figure 14: Convergence of FuRBO and SCBO on the 3D Spring design problem with three design variables, under four constraints. Results are averaged over 10 independent runs for each algorithm. The plot shows the mean objective value (volume), with shaded areas represent￾ing one s…
Figure 15
Figure 15. Figure 15: shows minimal difference between SCBO and FuRBO, which, as in the previous benchmark, can be attributed to the problem’s low dimensionality and weak constraint structure [PITH_FULL_IMAGE:figures/full_fig_p025_15.png]
Figure 16
Figure 16. Figure 16: Convergence of FuRBO and SCBO on the pressure vessel design problem with four design parameters and four constraints. Results are averaged over 10 independent runs for each algorithm. The plot shows the mean objective value (weight), with shaded areas representing one…
Figure 17
Figure 17. Figure 17: Convergence of FuRBO and SCBO on the speed reducer design problem with seven design parameters and eleven constraints. Results are averaged over 10 independent runs for each algorithm. The plot shows the mean objective value (volume), with shaded areas representing on…
Figure 18
Figure 18. Figure 18: Study on the trajectory-finding problem of a rover in dimension 60, under 15 constraints. (a) Visualization of the design domain problem and a trajectory solution. The yellow squares represent the obstacles the rover can overcome at a pre-fixed cost. The red squares r…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 20 canonical work pages

  1. [1]

    Amini, I

    S. Amini, I. Vannieuwenhuyse, and A. Morales-Hern \'a ndez . Constrained Bayesian Optimization : A Review . IEEE Access, 13: 0 1581--1593, 2025. ISSN 2169-3536. doi:10.1109/ACCESS.2024.3522876

    arXiv 2025

  2. [2]

    Ariafar, J

    S. Ariafar, J. Coll-Font , D. Brooks, and J. Dy. ADMMBO : Bayesian Optimization with Unknown Constraints using ADMM . Journal of Machine Learning Research, 20 0 (123): 0 1--26, 2019. ISSN 1533-7928

    work page 2019

  3. [3]

    Balandat, B

    M. Balandat, B. Karrer, D. R. Jiang, S. Daulton, B. Letham, A. G. Wilson, and E. Bakshy. BoTorch: A Framework for Efficient Monte-Carlo Bayesian Optimization . In Advances in Neural Information Processing Systems 33, 2020. URL http://arxiv.org/abs/1910.06403

    arXiv 2020

  4. [4]

    H. Cai, L. Zhu, and S. Han. ProxylessNAS : Direct Neural Architecture Search on Target Task and Hardware , Feb. 2019

    work page 2019

  5. [5]

    X. Du, J. Liang, J. Lei, J. Xu, and P. Xie. A radial-basis function mesh morphing and Bayesian Optimization framework for vehicle crashworthiness design. Structural and Multidisciplinary Optimization, 66 0 (3): 0 64, Mar. 2023. ISSN 1615-1488. doi:10.1007/s00158-023-03496-x

    doi:10.1007/s00158-023-03496-x 2023

  6. [6]

    Dufoss \'e and A

    P. Dufoss \'e and A. Atamna. Benchmarking several strategies to update the penalty parameters in AL-CMA-ES on the BBOB -constrained testbed. In Proceedings of the Genetic and Evolutionary Computation Conference Companion, pages 1691--1699, 2022

    work page 2022

  7. [7]

    Dufoss \'e , N

    P. Dufoss \'e , N. Hansen, D. Brockhoff, P. R. Sampaio, A. Atamna, and A. Auger. The BBOB-Constrained COCO Test Suite . 2022

    work page 2022

  8. [8]

    Eriksson and M

    D. Eriksson and M. Poloczek. Scalable constrained Bayesian optimization. In International conference on artificial intelligence and statistics, pages 730--738. PMLR, 2021. doi:10.48550/arXiv.2002.08526

    arXiv doi:10.48550/arxiv.2002.08526 2021

Show all 32 references
  1. [9]

    Eriksson, M

    D. Eriksson, M. Pearce, J. Gardner, R. D. Turner, and M. Poloczek. Scalable global optimization via local bayesian optimization. Advances in neural information processing systems, 32, 2019. doi:10.48550/arXiv.1910.01739. URL https://github.com/pytorch/botorch/blob/main/tutoria...

    2019 doi arXiv

  2. [10]

    Feurer, K

    M. Feurer, K. Eggensperger, S. Falkner, M. Lindauer, and F. Hutter. Auto- Sklearn 2.0: Hands-free AutoML via Meta-Learning , Oct. 2022

    2022

  3. [11]

    A. I. J. Forrester, A. S \'o bester, and A. J. Keane. Engineering Design via Surrogate Modelling - A Practical Guide . John Wiley & Sons Ltd., 2008. ISBN 978-0-470-06068-1

    2008

  4. [12]

    Gardner, M

    J. Gardner, M. Kusner, Zhixiang, K. Weinberger, and J. Cunningham. Bayesian Optimization with Inequality Constraints . In Proceedings of the 31st International Conference on Machine Learning , pages 937--945. PMLR, June 2014

    2014

  5. [13]

    R. Garnett. Bayesian Optimization . Cambridge University Press, 2023

    2023

  6. [14]

    N. Hansen. The CMA evolution strategy: a comparing review. Towards a new evolutionary computation: Advances in the estimation of distribution algorithms, pages 75--102, 2006. URL https://github.com/CMA-ES/pycma/blob/development/notebooks/notebook-usecases-constraints.ipynb

    2006

  7. [15]

    Hansen, Y

    N. Hansen, Y. Akimoto, and P. Baudis. CMA-ES/pycma on G ithub. Zenodo, Feb. 2019

    2019

  8. [16]

    Hansen, A

    N. Hansen, A. Auger, R. Ros, O. Mersmann, T. Tušar, and D. B. and. COCO: a platform for comparing continuous optimizers in a black-box setting . Optimization Methods and Software, 36 0 (1): 0 114--144, 2021. doi:10.1080/10556788.2020.1808977

    2021

  9. [17]

    J. M. Hern \'a ndez-Lobato , M. A. Gelbart, R. P. Adams, M. W. Hoffman, and Z. Ghahramani. A General Framework for Constrained Bayesian Optimization using Information-based Search , Sept. 2016

    2016

  10. [18]

    J. M. Hern \'a ndez-Lobato , J. Requeima, E. O. Pyzer-Knapp , and A. Aspuru-Guzik . Parallel and Distributed Thompson Sampling for Large-scale Accelerated Exploration of Chemical Space , June 2017

    2017

  11. [19]

    Hutter, L

    F. Hutter, L. Kotthoff, and J. Vanschoren. Automated Machine Learning: Methods, Systems, Challenges. Springer Nature, 2019

    2019

  12. [20]

    H. Jin, F. Chollet, Q. Song, and X. Hu. Autokeras: An automl library for deep learning. Journal of Machine Learning Research, 24 0 (6): 0 1--6, 2023. URL http://jmlr.org/papers/v24/20-1355.html

    2023

  13. [21]

    A. Keane. Experiences with optimizers in structural design. In Proceedings of the conference on adaptive computing in engineering design and control, volume 94, pages 14--27, 1994

    1994

  14. [22]

    A. C. Lemonge, H. J. Barbosa, C. C. Borges, and F. B. Silva. Constrained optimization problems in mechanical engineering design using a real-coded steady-state genetic algorithm. Mec \'a nica Computacional , 29 0 (95): 0 9287--9303, 2010

    2010

  15. [23]

    Picheny, R

    V. Picheny, R. B. Gramacy, S. M. Wild, and S. L. Digabel. Bayesian optimization under mixed constraints with a slack-variable augmented Lagrangian , May 2016

    2016

  16. [24]

    M. J. Powell. A direct search optimization method that models the objective and constraint functions by linear interpolation. Springer, 1994. URL https://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.minimize.html#scipy.optimize.minimize

    1994

  17. [25]

    W. B. Powell. A unified framework for stochastic optimization. European Journal of Operational Research, 275 0 (3): 0 795--821, June 2019. ISSN 03772217. doi:10.1016/j.ejor.2018.07.014

    2019 doi

  18. [26]

    Raponi, M

    E. Raponi, M. Bujny, M. Olhofer, N. Aulig, S. Boria, and F. Duddeck. Kriging-assisted topology optimization of crash structures. Computer Methods in Applied Mechanics and Engineering, 348: 0 730--752, May 2019. ISSN 0045-7825. doi:10.1016/j.cma.2019.02.002

    2019 doi

  19. [27]

    Schonlau, W

    M. Schonlau, W. J. Welch, and D. R. Jones. Global versus local search in constrained optimization of computer models. In New Developments and Applications in Experimental Design, volume 34, pages 11--26. Institute of Mathematical Statistics, 1998. doi:10.1214/lnms/1215456182

    1998

  20. [28]

    W. R. Thompson. On the Likelihood that One Unknown Probability Exceeds Another in View of the Evidence of Two Samples . Biometrika, 25 0 (3/4): 0 285--294, 1933. ISSN 0006-3444. doi:10.2307/2332286

    1933 doi

  21. [29]

    Ungredda and J

    J. Ungredda and J. Branke. Bayesian Optimisation for Constrained Problems . ACM Trans. Model. Comput. Simul., 34 0 (2): 0 9:1--9:26, Apr. 2024. ISSN 1049-3301. doi:10.1145/3641544

    2024 doi

  22. [30]

    Virtanen, R

    P. Virtanen, R. Gommers, T. E. Oliphant, M. Haberland, T. Reddy, D. Cournapeau, E. Burovski, P. Peterson, W. Weckesser, J. Bright, S. J. van der Walt , M. Brett, J. Wilson, K. J. Millman, N. Mayorov, A. R. J. Nelson, E. Jones, R. Kern, E. Larson, C. J. Carey, \.I . Polat, Y. F...

    2020

  23. [31]

    Z. Wang, C. Gehring, P. Kohli, and S. Jegelka. Batched large-scale bayesian optimization in high-dimensional spaces. In International Conference on Artificial Intelligence and Statistics, pages 745--754. PMLR, 2018. doi:10.48550/arXiv.1706.01445

    2018 doi arXiv

  24. [32]

    Zimmermann and O

    M. Zimmermann and O. De Weck. Formulating Engineering Systems Requirements . In A. Maier, J. Oehmen, and P. E. Vermaas, editors, Handbook of Engineering Systems Design , pages 1--52. Springer International Publishing , Cham , 2021. ISBN 978-3-030-46054-9. doi:10.1007/978-3-030...

    2021 doi

Pith tools

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