Pith. sign in

REVIEW 3 major objections 9 minor 1 cited by

Robust designs for Gaussian process emulation of computer experiments

T0 review · 3 major / 9 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper claims that support points—designs minimizing the energy distance between their empirical distribution and the uniform hypercube—are robust Gaussian-process emulation designs that perform near-optimally for both smooth and…

desk verdict A competent extension of the support-points framework to GP emulation, with useful identities and strong empirical evidence, but the central robustness claim rests on an unquantified surrogate approximation. read the letter →

arxiv 2507.09156 v1 pith:2JNBMSVN submitted 2025-07-12 stat.ME

Simon Mak , V. Roshan Joseph This is my paper
classification stat.ME MSC 62K0562K20
keywords computerexperimentsGaussianprocessemulationsupportpointsprojectedenergydistancerobustdesigndifference-of-convexprogrammingKoksma-Hlawkainequality
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 a design chosen to minimize the energy distance between its empirical distribution and the uniform distribution on the cube—the support points, and their projected variant—is a good default design for Gaussian process emulation. The claim is not that these designs are optimal for any one correlation model, but that they perform nearly optimally across a wide class of smooth and rugged response surfaces. The argument splits emulation into estimating the mean of the surface and estimating the variation around that mean. For the mean, a classical Koksma–Hlawka bound ties integration error to the energy distance; for the variation, an information-theoretic chain ties the sum of entries of a projected correlation matrix to the same energy distance under a mild Hölder-continuity assumption. The payoff is a design that avoids committing to correlation parameters in advance while matching or beating model-specific and distance-based competitors in the paper's numerical experiments.

What carries the argument

The central object is the energy distance $E(F,F_n) = 2\mathbb{E}\|X-Y\| - \mathbb{E}\|Y-Y'\| - \mathbb{E}\|X-X'\|$ between the target uniform distribution $F$ and the empirical distribution $F_n$ of a design. It simultaneously plays three roles: its minimization defines support points; it appears in the Koksma–Hlawka upper bound for the mean-estimation error; and, through Theorem 2, it serves as a surrogate for the maximum-entropy objective for variation estimation under the assumed correlation form. The projected support-point objective replaces the Euclidean norm with a generalized Gaussian kernel weighted by a prior on scale parameters, encoding effect sparsity, hierarchy, and heredity.

What would settle it

For a fixed small design in two dimensions, compute the true maximum-entropy design (maximizing $\det R_n$) for a Gaussian correlation with moderate correlation length and compare its integrated prediction error with that of the support-point design across a grid of test functions satisfying the Hölder condition. If support points are consistently and substantially worse than the true maximum-entropy design, the surrogate-objective step is not doing the claimed work. A second check is whether the lower bound in equation (12) is tight: if the design minimizing the sum of correlation entries does not track the design maximizing $\det R_n$ as $n$ grows, the approximation fails.

Watch

Extended reading notes

Core claim

Under the correlation model $r_\theta(x,y) \propto \|x\|^2 + \|y\|^2 - \|x-y\|^2$, the entry-wise sum of the projected correlation matrix equals, up to a constant, the energy distance $E(F,F_n)$ between the uniform distribution $F$ and the empirical distribution of the design. Since minimizing the energy distance is exactly the support-point criterion, support points maximize information about the variation term of a Gaussian process while simultaneously optimizing a tight Koksma–Hlawka bound for the mean term, assuming integrable derivatives. The paper further shows that support points approximate a two-parts-minimax, one-part-maximin compromise, and that projected support points under an exponential sparsity prior recover the MaxPro criterion plus a correction that pulls points away from the boundary.

Load-bearing premise

The paper replaces the maximum-entropy design objective with a lower bound, assuming that minimizing the sum of all correlation-matrix entries closely approximates maximizing the determinant, but no tightness bound for this lower bound is proved.

Editorial extensions

If this is right

  • Support points and projected support points can be generated efficiently by difference-of-convex programming, so large designs in high dimensions become computationally practical.
  • In the paper's numerical suite, support points and projected support points achieved the best efficiency in nearly every smooth-versus-rough and all-versus-some-active-factors combination, with projected support points especially strong when only some factors are active.
  • Support points provide an explicit trade-off between minimax and maximin behavior, with roughly double weight on the minimax (fill-distance) component, jointly addressing nominal and numeric interpolation errors.
  • Projected support points recover MaxPro designs with an added boundary-repulsion correction, improving projected-space uniformity and emulation for functions with sparse active effects.
  • The connection to fractional Brownian motion with Hurst index $1/2$ suggests the design supports non-stationary emulation where predictions do not revert to the mean.

Reading between the lines

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

  • Because the theoretical argument depends only on a Hölder-type continuity condition rather than specific correlation values, the same energy-distance design should transfer to Gaussian process models with estimated or unknown correlation lengths; this is a natural extension the paper does not prove.
  • The two-parts-minimax, one-part-maximin weighting is derived in a small-$q$ limit, so whether a different weighting depending on design size or dimension would improve emulation remains open and is testable.
  • Minimizing the energy distance is equivalent to minimizing a negative-definite kernel discrepancy, so support points inherit quasi-Monte Carlo integration guarantees for the mean term and may serve double duty as an integration rule in the same experiment.
  • For non-uniform target distributions $F$, the same construction should yield designs weighted toward regions where prediction accuracy matters most, a direction the paper notes but does not test.
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

3 major / 9 minor

Summary. The paper studies two classes of experimental designs, support points (SPs) and projected support points (PSPs), for Gaussian-process emulation of computer experiments. It argues that SPs are robust designs because they provide (i) a tight Koksma–Hlawka bound on the integration error for the mean term of the response, and (ii) near-maximal Shannon information on the variation term via a maximum-entropy surrogate argument involving the energy distance. The paper also derives connections between SPs and minimax/maximin designs, connects PSPs to MaxPro designs through a closed-form objective under an exponential prior, proposes algorithmic improvements (informed initialization and randomized Sobol' batch sampling), and reports numerical experiments on distance metrics, an IRMSE-based efficiency criterion, and four test functions. The theoretical sections contain several correct and useful identities, but the central robustness claim hinges on an unquantified surrogate-objective approximation (eqs. (11)–(13)) and the small-q approximation in Proposition 2 has uncontrolled errors.

Significance. If the central robustness claim holds, the paper offers a practically attractive answer to a well-known problem: GP emulation designs are typically sensitive to the assumed correlation parameters, whereas SPs/PSPs are cheap to generate in high dimensions and empirically competitive across smooth and rough surfaces. The manuscript contains several correct and useful calculations: Lemma 1 is a standard projection-kernel identity and is proved cleanly; Theorem 3 (integration error of fBm sample paths equals q-energy distance) is a correct and elegant identity; Theorem 4 gives a genuinely informative closed-form connection between PSPs and MaxPro designs with a boundary-correction term; and the numerical study, including the efficiency metric Eff(D;Theta) and the four test functions, is a strength because the comparisons are measured rather than fitted. The algorithmic claims are supported by the cited prior works and the new initialization/Sobol-batch suggestions are motivated, though their added benefit is reported only informally.

major comments (3)
  1. [Section 3.2, eqs. (11)–(13)] The central robustness claim that SPs provide 'near-maximal information gain' on P_perp_G f relies on replacing the maximum-entropy objective det(R_n) by the sum of correlation entries sum_{i,j} r_theta(x_i,x_j). The text states that this follows from Gershgorin's circle theorem, but Gershgorin only gives a lower bound on the determinant in terms of row sums; maximizing that lower bound is not equivalent to maximizing det(R_n) unless tightness or order-preservation over the design space is established, and no such argument is given. The step from (12) to (13) additionally replaces the objective by the lower bound 2nr - sum r_ij, and maximizing a lower bound is not equivalent to maximizing the original quantity when the bound is loose. A concrete failure mode exists even at n = 3: for a fixed sum of off-diagonal correlations, det(R) depends on the configuration (equal pairwise correlations give a larger determinant than one large and two small correlations with the same sum), so minimizing the sum cannot distinguish designs with different entropies. Because Theorem 2 only proves an exact identity for a specific kernel after the surrogate has been adopted, the chain (11)->(13) is the load-bearing step for the information-theoretic interpretation. I recommend either proving a tightness or approximation bound for the surrogate, or reframing the theoretical claim as a heuristic motivation and relying on the numerical evidence for robustness.
  2. [Section 4.2, Proposition 2 and eqs. (19)–(21)] The claim that SPs are 'two-parts minimax and one-part maximin' rests on approximations (19) and (20), whose errors are uncontrolled and may be of the same order as the quantities being estimated as q -> 0+. For finite q, the right-hand side of (21) approximates E_q(F,F_n) with an error that depends on the variation of h^q over Voronoi regions and on the gap between the q-th power mean and the minimum over pairs; no bound is supplied, and the approximation is used to draw a design-principle conclusion (greater weight on minimax than maximin). This is not a fatal flaw for the paper's practical claim, but the statement 'SPs can be viewed as designs which are two-parts minimax and one-part maximin' in Section 4.2 should be softened or accompanied by a quantitative error analysis, since it is used in the discussion of nominal versus numeric error control.
  3. [Section 3.2, Theorem 2 and Table 1] The assumption r_theta(x,y) proportional to ||x||+||y||-||x-y|| is a very specific correlation structure (an fBm(1/2)-type kernel), and Theorem 2 is stated only for this kernel. The robustness claim in Table 1, however, is phrased broadly for the variation term under 'expected 1/2-Holder continuity'. The passage from the surrogate approximation (13) with a general kernel r_theta to the energy-distance identity (14) for this one kernel is not justified for other kernels, so the breadth of the claim as summarized in Table 1 is larger than what the proof establishes. Please clarify the role of the specific kernel in the theorem and state more precisely which kernels the near-maximal-information claim covers.
minor comments (9)
  1. [Section 3.2, eq. (11)] Equation (11) is introduced as an 'approximate formulation' citing Section 5 of [18], but the connection to Gershgorin's theorem is not derived in the text; a short derivation or a more precise citation (equation number in [18]) would help the reader assess the approximation.
  2. [Section 2.1, Definition 2] The notation in the proof of Theorem 2 sums over 'j=i' in the first expression of eq. (14), whereas the intended sum appears to be over all i,j = 1,...,n; this appears to be a typo that should be corrected.
  3. [Section 4.2, eq. (23)] The approximation of the average of ||x_i - x_j||^q by the minimum [M_m(D)]^q requires justification beyond eq. (20), since the minimum is not a representative value of the distribution of pairwise distances; at minimum, the statement 'sufficiently small q' should be accompanied by an explicit error term.
  4. [Section 4, Proposition 2 proof] In the proof, the use of eq. (20) to replace the average of 2/n sum_i ||z_i - x_i||^q by 2 max_i ||z_i - x_i||^q is not obvious, since the maximum is not the average of the values; this step should be clarified or corrected.
  5. [Section 5.2, Theorem 4] The statement 'PSPs correct the boundary attraction behavior of MaxPro designs' is supported for the exponential-prior anisotropic kernel, but the PSPs used in the simulations use the POD prior (29); the transfer of the theoretical boundary-correction insight to the simulation setting should be stated as an interpretation, not as a proven property of the POD-based PSPs.
  6. [Section 6 and Figure 3] The timing comparison in Figure 3 is reported without a description of the software/hardware setup for the other designs (only the processor is mentioned for SPs/PSPs); adding the implementation details for minimax, MmLHD, MaxPro, and UD would make the computational comparison reproducible.
  7. [Section 7, Table 2] The efficiency metric Eff(D;Theta) uses D*(theta) estimated as the lowest IRMSE among the six designs, which makes the reported efficiencies relative to a finite set of designs rather than the true optimum; this is fine for ranking the six designs, but the phrase 'D*(theta)' in eq. (31) should be defined consistently as the best design among those considered, not as the global argmin.
  8. [Section 8, Table 4] The fEff values are reported only as averages over n for each function; reporting standard errors or per-n full values (beyond the underline convention) would allow the reader to assess the variability of the comparisons.
  9. [Throughout] There are a few typographical issues, e.g., 'absense' in Section 7.3.1, 'Alterative' in Table 3, and the incomplete sentence in the proof of Theorem 2 ('this completes the proof' following an incorrect sum index); a careful proofreading pass is recommended.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's support-point optimality claims follow from proven energy-distance identities and measured comparisons, not from a fitted parameter or a self-citation chain.

full rationale

I walked the claimed derivation chain. Support points are definitionally minimizers of the energy distance E(F,Fn) (Definition 2), and the paper's theoretical results are substantive identities connecting that defining objective to GP emulation quantities: Lemma 1 derives the projected kernel rθ,G; Proposition 1 is a Koksma-Hlawka bound; Theorem 2 proves that a sum of projected-kernel entries is proportional to E(F,Fn); Theorem 3 proves that fBm integration error equals a q-energy distance; Proposition 2 and Theorem 4 give further exact or explicitly approximate connections. None of these steps inverts an input into an output or fits a parameter and then renames it as a prediction. The weakest link is Section 3.2's chain (11)-(13), where maximum entropy designs are approximated by minimizers of the sum of correlation entries. That approximation is asserted without a tightness or error bound, so the 'near-maximal information' claim is not fully rigorous. However, this is a correctness/rigor gap, not circularity: maximum entropy is an external criterion, the paper does not define it as energy distance, and the approximation is stated openly rather than smuggled in. Self-citations to [25], [27], and [18] provide previously published definitions, convergence results, and algorithms that do not assume the paper's conclusion; they are independent support under the stated rules. The empirical efficiency comparisons are measured against independently generated designs, and the best-of-six oracle used to estimate D*(θ) is a benchmark choice rather than a fitted input. No step reduces, by construction or by self-citation, to its own inputs.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central claims rest on the energy-distance framework, the Sobolev and Holder regularity assumptions on the response surface, and two heuristic approximations: the surrogate maximum-entropy objective and the small-q approximation. No new entities are introduced. The designs themselves come from the authors' prior work.

free parameters (1)
  • POD prior hyperparameters for PSPs = theta_l ~ Gamma(0.1, 1); Gamma_{|u|} = p^{-1/4}(|u|!)^{-1/2}
    The projected support point objective (4) requires a prior pi on the kernel scale parameters. The POD specification (29) is chosen by hand to encode effect hierarchy, heredity and sparsity; different hyperparameters would produce different designs, and no sensitivity analysis is reported.
assumptions (5)
  • domain assumption Expected 1/2-Holder continuity E[{f(x)-f(y)}^2] = O(||x-y||^2)
    Invoked in Section 3.2 before Theorem 2 to justify the Brownian covariance kernel ||x||^2+||y||^2-||x-y||^2 as the relevant correlation model; restricts the response surface class.
  • ad hoc to paper The Gershgorin lower bound (12) is tight enough that maximizing it approximates maximum entropy designs (11)
    Section 3.2, equations (11)-(13): the paper replaces the max-entropy objective with the sum of correlation entries using a lower bound, without error control. This is the main load-bearing heuristic for the information-gain claim.
  • ad hoc to paper Small-q approximations (19)-(20) hold with negligible error relative to E_q(F,F_n)
    Section 4.2, Proposition 2: the paper approximates h^q(z) by its average and the pair-sum by the minimum distance for small q, but the residual errors are of the same order as the q-energy distance being approximated.
  • standard math Bounded convergence theorem for interchange of integrals and expectations
    Used in the proofs of Lemma 1 and Theorem 3 on the bounded domain [0,1]^p.
  • domain assumption Sobolev regularity f in W^{ceil((p+1)/2),2} for the Koksma-Hlawka bound
    Proposition 1, equation (10), requires the integrand's ceil((p+1)/2)-th order weak derivatives to be square-integrable.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust designs for Gaussian process emulation of computer experiments." pith.science (2026). https://pith.science/paper/2JNBMSVN

@misc{pith2026250709156,
  author       = {Pith},
  title        = {Pith review of: Robust designs for Gaussian process emulation of computer experiments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2JNBMSVN}},
  note         = {Machine review of arXiv:2507.09156}
}
read the original abstract

We study in this paper two classes of experimental designs, support points and projected support points, which can provide robust and effective emulation of computer experiments with Gaussian processes. These designs have two important properties that are appealing for surrogate modeling of computer experiments. First, the proposed designs are robust: they enjoy good emulation performance over a wide class of smooth and rugged response surfaces. Second, they can be efficiently generated for large designs in high dimensions using difference-of-convex programming. In this work, we present a theoretical framework that investigates the above properties, then demonstrate their effectiveness for Gaussian process emulation in a suite of numerical experiments.

Figures

Figures reproduced from arXiv: 2507.09156 by the authors.

Figure 1
Figure 1. GP fits of the scaled Runge function and the modified step function using n = 7 SPs and Chebyshev nodes. Solid lines indicate true functions, dotted lines indicate predictions, dots indicate observed function points, and MSPE indicates mean-squared prediction error. 3.3. Support points and fractional Brownian motion modeling The correlation function in Theorem 2 also has a useful connection to a special class of GPs… view at source ↗
Figure 2
Figure 2. A visualization of the correction terms ϕ(x; 1) and ϕ(x; 0.5). The maxi￾mization of this term pushes design points away from the boundaries of [0, 1]. In other words, subject to the sparsity prior θl i.i.d. ∼ Exp(λ), the resulting PSPs jointly minimize the MaxPro criterion in (27), and maximize an expression involving the correction term ϕ(x; λ). For illustration, [PITH_FULL_IMAGE:figures/full_fig_p022_2.png] view at source ↗
Figure 3
Figure 3. Computation time (in seconds) for n = 10p-point designs on [0, 1]p , for various dimensions p. such designs to provide an informed initialization of sp.sccp or psp.sccp. Put another way, sp.sccp or psp.sccp can be viewed as a post-processing step which improves space-filling properties of an existing design, all the while retaining desired characteristics of this original design. As indicated in Algorithms 1 and 2, … view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: n = 25-point designs for SPs and PSPs on F = U[0, 1]2 , and the minimax, MmLHD, uniform and MaxPro designs. (Left) 2-d visualization; (Right) 1-d projections [PITH_FULL_IMAGE:figures/full_fig_p036_4.png]
Figure 5
Figure 5. Figure 5: (Top) Minimax and maximin index criteria for n = 75-point SPs on U[0, 1]10, MmLHDs and minimax designs. (Bottom) Minimax and maximin index criterion for n = 75-point PSPs on U[0, 1]10, MaxPro and miniMaxPro designs. Smoothness Θ = (0, 5]s Θ = (0, 20]s % active 100% act…

Discussion (0). Sign in 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. Neural Low-Discrepancy Sequences

    cs.LG 2025-10 conditional novelty 6.0 of 10

    A trained index-to-point network generates low-discrepancy sequences whose prefixes beat Sobol, Halton, and scrambled Sobol discrepancy in four dimensions.

Reference graph

Works this paper leans on

55 extracted references · 53 canonical work pages · cited by 1 Pith paper

  1. [1]

    Ba, S. (2015). SLHD: Maximin-Distance (Sliced) Latin Hypercube Designs . R package version 2.1.1

    work page 2015

  2. [2]

    and Joseph, V

    Ba, S. and Joseph, V. R. (2015). MaxPro: Maximum Projection Designs . R package version 3.1.2

    work page 2015

  3. [3]

    Bernardini, M., Cimini, M., Stella, F., Cavallini, E., Di Mascio, A., Neri, A., and Martelli, E. (2020). Large-eddy simulation of vortex shedding and pressure oscillations in solid rocket motors. AIAA Journal, 58(12):5191– 5201

    work page 2020

  4. [4]

    Box, G. E. and Hunter, J. S. (1961). The 2 k−p fractional factorial designs. Technometrics, 3(3):311–351

    work page 1961

  5. [5]

    J., and Yang, V

    Chang, Y.-H., Wang, X., Zhang, L., Li, Y., Mak, S., Wu, C.-F. J., and Yang, V. (2021). Reduced-order modeling for complex flow emulation by common kernel-smoothed proper orthogonal decomposition. AIAA Journal, 59(9):3291–3303

    work page 2021

  6. [6]

    and Pepelyshev, A

    Dette, H. and Pepelyshev, A. (2010). Generalized Latin hypercube design for computer experiments. Technometrics, 52(4):421–429

    work page 2010

  7. [7]

    Y., and Sloan, I

    Dick, J., Kuo, F. Y., and Sloan, I. H. (2013). High-dimensional integration: the quasi-Monte Carlo way. Acta Numerica, 22:133–288

    work page 2013

  8. [8]

    Fang, K.-T. (1980). The uniform design: application of number-theoretic methods in experimental design. Acta Math. Appl. Sinica , 3(4):363–372

    work page 1980

Show all 55 references
  1. [9]

    Forrester, A., Sobester, A., and Keane, A. (2008). Engineering Design via Surrogate Modelling: A Practical Guide . John Wiley & Sons

    2008

  2. [10]

    H., Grosse, E., and Stuetzle, W

    Friedman, J. H., Grosse, E., and Stuetzle, W. (1983). Multidimensional additive spline approximation. SIAM Journal on Scientific and Statistical Computing, 4(2):291–301

    1983

  3. [11]

    Gershgorin, S. A. (1931). Uber die abgrenzung der eigenwerte einer matrix. Izv. Akad. Nauk. USSR Otd. Fiz.-Mat. Nauk , (6):749–754

    1931

  4. [12]

    and Qian, P

    Haaland, B. and Qian, P. Z. (2011). Accurate emulators for large-scale computer experiments. The Annals of Statistics , 39(6):2974–3002

    2011

  5. [13]

    and Wu, C

    Hamada, M. and Wu, C. F. J. (1992). Analysis of designed experiments with complex aliasing. Journal of Quality Technology , 24(3):130–137

    1992

  6. [14]

    He, X. (2017). Interleaved lattice-based minimax distance designs. Biometrika, 104(3):713–725. S. Mak and V. R. Joseph/Robust designs for computer experiments 33

    2017

  7. [15]

    Hickernell, F. (1998). A generalized discrepancy and quadrature error bound. Mathematics of Computation of the American Mathematical Society, 67(221):299–322

    1998

  8. [16]

    S., Soeder, D., Paquet, J.-F., Bass, S

    Ji, Y., Yuchi, H. S., Soeder, D., Paquet, J.-F., Bass, S. A., Joseph, V. R., Wu, C. F., and Mak, S. (2024). Conglomerate multi-fidelity Gaussian process modeling, with application to heavy-ion collisions. SIAM/ASA Journal of Uncertainty Quantification , 12:473–502

    2024

  9. [17]

    E., Moore, L

    Johnson, M. E., Moore, L. M., and Ylvisaker, D. (1990). Minimax and maximin distance designs. Journal of Statistical Planning and Inference , 26(2):131–148

    1990

  10. [18]

    R., Gul, E., and Ba, S

    Joseph, V. R., Gul, E., and Ba, S. (2015). Maximum projection designs for computer experiments. Biometrika, 102(2):371–380

    2015

  11. [19]

    Kolmogorov, A. (1933). Sulla determinazione empirica delle leggi di probabilita. Giorn. Ist. Ital. Attuari , 4:1–11

    1933

  12. [20]

    Li, K., Mak, S., Paquet, J.-F., and Bass, S. A. (2025). Additive multi- index gaussian process modeling, with application to multi-physics surrogate modeling of the quark-gluon plasma. Journal of the American Statistical Association, (just-accepted)

    2025

  13. [21]

    L., Sacks, J., and Welch, W

    Loeppky, J. L., Sacks, J., and Welch, W. J. (2009). Choosing the sample size of a computer experiment: A practical guide. Technometrics, 51(4):366– 376

    2009

  14. [22]

    MacDonald, B., Ranjan, P., and Chipman, H. (2015). GPfit: an R package for fitting a Gaussian process model to deterministic simulator outputs. Journal of Statistical Software , 64(12)

    2015

  15. [23]

    Mak, S. (2016). minimaxdesign: Minimax and Minimax Projection Designs. R package version 0.1.2

    2016

  16. [24]

    Mak, S. (2017). support: Support Points . R package version 0.1.0

    2017

  17. [25]

    and Joseph, V

    Mak, S. and Joseph, V. R. (2017). Projected support points, with application to optimal MCMC reduction. arXiv preprint arXiv:1708.06897

    2017 arXiv

  18. [26]

    and Joseph, V

    Mak, S. and Joseph, V. R. (2018). Minimax and minimax projection designs using clustering. Journal of Computational and Graphical Statistics , 27(1):166–178

    2018

  19. [27]

    Mak, S and Joseph, V. R. (2018). Support points. Ann Statist , 46(6A):2562–2592

    2018

  20. [28]

    R., Yang, V., and Wu, C

    Mak, S, Sung, C.-L., Wang, X., Yeh, S.-T., Chang, Y.-H., Joseph, V. R., Yang, V., and Wu, C. F. J. (2018). An efficient surrogate model for emulation and physics extraction of large eddy simulations. J Amer Statist Assoc, 113(524):1443–1456. S. Mak and V. R. Joseph/Robust desi...

    2018

  21. [29]

    Mandelbrot, B. B. and Van Ness, J. W. (1968). Fractional Brownian motions, fractional noises and applications. SIAM Review, 10(4):422–437

    1968

  22. [30]

    J., Mak, S., Sun, B., Narayanan, S

    Miller, J. J., Mak, S., Sun, B., Narayanan, S. R., Yang, S., Sun, Z., Kim, K. S., and Kweon, C.-B. M. (2024). Expected Diverse Improvement (EDU): Diverse Bayesian optimization of expensive computer simulators. arXiv preprint arXiv:2410.01196

    2024 arXiv

  23. [31]

    Morris, M. D. and Mitchell, T. J. (1995). Exploratory designs for computational experiments. Journal of Statistical Planning and Inference , 43(3):381–402

    1995

  24. [32]

    R., Ji, Y., Sapra, H

    Narayanan, S. R., Ji, Y., Sapra, H. D., Kweon, C.-B. M., Kim, K. S., Sun, Z., Kokjohn, S., Mak, S., and Yang, S. (2024). A misfire-integrated Gaussian process (MInt-GP) emulator for energy-assisted compression ignition (EACI) engines with varying cetane number jet fuels. Inter...

    2024

  25. [33]

    Nualart, D. (2006). Fractional Brownian motion: stochastic calculus and applications. In International Congress of Mathematicians , volume 3, pages 1541–1562

    2006

  26. [34]

    Owen, A. B. (1998). Scrambling Sobol’ and Niederreiter–Xing points. Journal of Complexity , 14(4):466–489

    1998

  27. [35]

    and Joseph, V

    Plumlee, M. and Joseph, V. R. (2018). Orthogonal Gaussian process models. Statistica Sinica, pages 601–619

    2018

  28. [36]

    T., Harari, O., Bingham, D., and Flowers, G

    Pratola, M. T., Harari, O., Bingham, D., and Flowers, G. E. (2017). Design and analysis of experiments on nonconvex regions. Technometrics, 59(1):36–47

    2017

  29. [37]

    Pukelsheim, F. (2006). Optimal Design of Experiments . SIAM

    2006

  30. [38]

    Putschke, J., Kauder, K., Khalaj, E., Angerami, A., Bass, S., Cao, S., Coleman, J., Cunqueiro, L., Dai, T., Du, L., et al. (2019). The JETSCAPE framework. arXiv preprint arXiv:1903.07706

    2019 arXiv

  31. [39]

    B., and Welch, W

    Sacks, J., Schiller, S. B., and Welch, W. J. (1989a). Designs for computer experiments. Technometrics, 31(1):41–47

    1989

  32. [40]

    J., Mitchell, T

    Sacks, J., Welch, W. J., Mitchell, T. J., and Wynn, H. P. (1989b). Design and analysis of computer experiments. Statistical Science, 4(4):409–423

    1989

  33. [41]

    J., Williams, B

    Santner, T. J., Williams, B. J., and Notz, W. I. (2013). The Design and Analysis of Computer Experiments . Springer Science & Business Media

    2013

  34. [42]

    Discovering JMP 13

    SAS Institute (2016). Discovering JMP 13 . SAS Institute

    2016

  35. [43]

    Shannon, C. E. (1948). A mathematical theory of communication. Bell System Technical Journal, 27:623–656

    1948

  36. [44]

    Shewry, M. C. and Wynn, H. P. (1987). Maximum entropy sampling. S. Mak and V. R. Joseph/Robust designs for computer experiments 35 Journal of Applied Statistics , 14(2):165–170

    1987

  37. [45]

    Sz´ ekely, G. J. and Rizzo, M. L. (2004). Testing for equal distributions in high dimension. InterStat, 5:1–6

    2004

  38. [46]

    Sz´ ekely, G. J. and Rizzo, M. L. (2013). Energy statistics: A class of statistics based on distances. Journal of Statistical Planning and Inference , 143(8):1249–1272

    2013

  39. [47]

    Tan, M. H. (2013). Minimax designs for finite design regions. Techno- metrics, 55(3):346–358

    2013

  40. [48]

    Trefethen, L. N. (2013). Approximation Theory and Approximation Practice. SIAM

    2013

  41. [49]

    Tuo, R. (2017). Adjustments to computer models via projected kernel calibration. arXiv preprint arXiv:1705.03422

    2017 arXiv

  42. [50]

    Wendland, H. (2004). Scattered Data Approximation. Cambridge Uni- versity Press

    2004

  43. [51]

    Wiens, D. P. (1991). Designs for approximately linear regression: two optimality properties of uniform designs. Statistics & Probability Letters , 12(3):217–221

    1991

  44. [52]

    Wiens, D. P. and Zhou, J. (1997). Robust designs based on the in- finitesimal approach. Journal of the American Statistical Association , 92(440):1503–1511

    1997

  45. [53]

    Wu, C. F. J. and Hamada, M. S. (2009).Experiments: Planning, Analysis, and Optimization . John Wiley & Sons

    2009

  46. [54]

    Yuille, A. L. and Rangarajan, A. (2003). The concave-convex procedure. Neural Computation, 15(4):915–936

    2003

  47. [55]

    and Apley, D

    Zhang, N. and Apley, D. W. (2014). Fractional Brownian fields for response surface metamodeling. Journal of Quality Technology , 46(4):285– 301. S. Mak and V. R. Joseph/Robust designs for computer experiments 36 Figure 4: n = 25 -point designs for SPs and PSPs on F = U [0, 1]2...

    2014

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.