Pith. sign in

REVIEW 3 major objections 4 minor 46 references

Adaptive Resolution for Finite-Rank Gaussian Processes

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Putting a prior on grid resolution restores Gaussian-process optimality.

desk verdict The d=1 story is credible, but the multivariate version of Theorem 3.5 does not follow from the proof as written: the displayed exponents drop the d-dependence. read the letter →

arxiv 2505.24066 v2 pith:VH34NPMO submitted 2025-05-29 math.ST stat.MEstat.MLstat.TH

Jaehoan Kim , Anirban Bhattacharya , Debdeep Pati This is my paper · ORCID
classification math.STstat.MEstat.MLstat.TH MSC 62G2062F1560G15
keywords Gaussianprocessposteriorcontractionadaptiveinferenceminimaxratefinite-rankapproximationSPDEgridinterpolationhierarchicalprior
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 studies cheap approximations to Gaussian-process (GP) regression built from finitely many locally supported basis functions with dependent Gaussian coefficients, and asks whether such approximations can match the statistical efficiency of the full GP. For a grid-interpolation construction with a squared-exponential kernel, the authors prove that placing a hierarchical prior on the resolution $N$ and the inverse bandwidth $\kappa$ yields the minimax-optimal posterior contraction rate $n^{-\alpha/(2\alpha+d)}$ up to logarithmic factors for every $\alpha$-Hölder truth, without the user knowing $\alpha$. For the SPDE-based finite-element construction with a fixed Matérn smoothness $\beta$, they prove the opposite: the posterior concentrates at a strictly slower rate for sufficiently smooth functions regardless of bandwidth and truncation, so it cannot adapt to global smoothness beyond a threshold. The positive result matters because it indicates that a computationally cheap lattice interpolation can inherit the full GP's adaptive efficiency, provided the resolution itself is learned rather than fixed.

What carries the argument

The load-bearing object is the grid-interpolation GP $f_N(x)=\sum_{0\le j\le N} w_j\psi_j(x)$, where $\psi_j$ are piecewise-linear tent functions on a regular lattice and the coefficient vector $(w_0,\dots,w_N)$ is drawn from a parent squared-exponential GP evaluated at the lattice points, so the prior covariance is a Toeplitz or block-Toeplitz matrix. The argument runs through the GP concentration function: the proof lower-bounds the finite-rank prior's small-ball probability $P(\|f^* - f_N\|_\infty < \epsilon)$ by the full GP's small-ball probability $P(\|f^* - f\|_\infty < \epsilon/2)$ minus a fluctuation term $P(\|f - f_N\|_\infty > \epsilon/2)$, which is controlled by a sup-norm continuity estimate for the RBF GP over cells of width $1/N$. A polynomial-tailed prior on $N$ guarantees that with high probability there are enough grid points to make the interpolation error negligible at the relevant $\epsilon$-scale, while the gamma-type prior on $\kappa$ supplies the optimal concentration of the parent GP. The SPDE construction is analyzed by the same small-ball calculus, but there the fixed Matérn smoothness $\beta$ caps the effective regularity and produces the slower lower-bound rate.

What would settle it

Run the grid-interpolation posterior with a squared-exponential kernel on data from a very smooth function using a prior on $N$ with $P_2(N \ge x) = K_1 x^{-K_2}$, and measure the empirical $L^2$ error at growing sample sizes; if the error does not decay at essentially $n^{-\alpha/(2\alpha+d)}$, Theorem 3.5 is false. Conversely, simulating with an exponentially tailed prior on $N$ and observing a slower rate would indicate condition (12) is not merely technical.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3.5: for $f^* \in C^\alpha[0,1]^d$, if $f_N$ is the regular-grid linear interpolation of a squared-exponential GP with kernel $K(x)=\exp(-\kappa^2\|x\|^2)$, and independent priors are placed on $\kappa$ satisfying a gamma-type tail condition and on $N$ satisfying a polynomial-tail condition $P_2(N \ge x) \ge K_1 x^{-K_2}$, then the posterior contracts at $\epsilon_n = n^{-\alpha/(2\alpha+d)}(\log n)^C$ in the empirical $L^2$ metric. This is the same minimax-optimal rate as the unapproximated GP with inverse-gamma bandwidth, obtained adaptively over all $\alpha$. The mechanism is that for $N$ large enough relative to $\kappa$, the interpolation error $f_N - f$ is of higher order than the GP's own concentration scale, so the finite-rank posterior inherits the full GP's prior concentration around $f^*$. The paper's contrasting result, Theorem 3.4, shows that the SPDE finite-element prior with Matérn smoothness $\beta=2$ or even $\beta \ge 4$ concentrates at best at $n^{-(\beta-1/2)/(2\beta)}(\log n)^{1/4}$, which is suboptimal for truths smoother than $\beta - 1/2$; for $\beta=2$ the lower bound is $(n/\log n)^{-2/5}$. The paper also provides a Metropolis-within-Gibbs sampler that jointly updates $N$ and $\kappa$ for the interpolation model, with per-iteration cost $O(n + N^{3d})$, and numerical studies in $d=1$ and $d=2$.

Load-bearing premise

The guarantee of adaptivity rests on condition (12), that the prior on the resolution $N$ has at least a polynomial tail; the paper's own simulations use bounded or exponential priors on $N$, so those numerics do not fall within the theorem's assumptions and cannot directly confirm the rate.

Editorial extensions

If this is right

  • With a squared-exponential parent kernel, the grid-interpolation GP with hierarchical priors on $\kappa$ and $N$ attains the minimax-optimal posterior contraction rate $n^{-\alpha/(2\alpha+d)}(\log n)^C$ adaptively for every $\alpha>0$, so no oracle knowledge of smoothness is needed.
  • The SPDE finite-element approximation with fixed Matérn smoothness $\beta$ cannot adapt to functions smoother than $\beta-1/2$: its contraction rate is suboptimal by a polynomial factor for such truths.
  • The hierarchical sampler updates resolution and bandwidth together at per-iteration cost $O(n + N^{3d})$, so the theoretical guarantees come with a tractable computational scheme whose cost is dominated by the resolution rather than the sample size.
  • Because posterior contraction at rate $\epsilon_n$ implies existence of a point estimator with the same rate, the result supplies a practical estimation procedure that is both scalable and rate-optimal.

Reading between the lines

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

  • The theorem's tail condition on $N$ is at least polynomial, but the simulations use bounded or exponential priors on $N$; extending the proof to sub-polynomial tails would be needed before the numerics sit inside the theorem's scope.
  • The same interpolation-over-resolution mechanism might transfer to other covariance kernels with known small-ball probabilities, not just the squared-exponential kernel, since the fluctuation bound only needs a modulus-of-continuity control on the parent GP.
  • A practical reading is that practitioners should favor heavy-tailed or at-least-polynomial priors on the resolution when using grid GP emulators, because the theory ties adaptivity to that tail; discrete uniform or exponential priors may still work but are currently unsupported.
  • The suboptimality of the SPDE construction suggests that letting the Matérn smoothness itself be random, rather than only $\kappa$ and $N$, would be a natural next step; the paper does not pursue that direction.
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 / 4 minor

Summary. The paper studies finite-rank Gaussian process priors of the form f_N = sum_j w_j psi_j with dependent Gaussian coefficients and locally supported basis functions, focusing on two constructions: a finite-element/SPDE approximation of Matérn GPs and a regular-grid interpolation (GPI) prior induced by a parent GP. The main theoretical claims are that the SPDE construction has a suboptimal posterior contraction rate for sufficiently smooth true functions satisfying a uniform nonlinearity condition, while the GPI construction, equipped with independent hierarchical priors on the inverse bandwidth and the resolution parameter N, attains the adaptive minimax rate n^{-alpha/(2alpha+d)} up to logarithmic factors over alpha-Hölder classes. The paper also develops a Metropolis-within-Gibbs sampler for the hierarchical GPI model and reports simulations comparing the methods with full GP regression.

Significance. If the main results are correct, the paper makes a useful contribution: it shows that a finite-rank interpolation prior can inherit the adaptive minimax behavior of the full Gaussian process while reducing computational cost, and it provides quantitative evidence of the limitations of finite-element SPDE approximations. The paper contains substantial technical work, including explicit small-ball probability bounds for the SPDE prior, fluctuation bounds for interpolated Gaussian processes, and a concrete sampling scheme with Toeplitz/BTTB structure. The central rate result is built on external benchmarks from van der Vaart and van Zanten rather than on a circular use of the authors' own fitted quantities. However, the d-dimensional proof of Theorem 3.5 has a dimensional gap that currently limits the theorem to d=1, and the statement of Theorem 3.4 appears inconsistent with its proof; these issues need to be repaired before the central claims can be accepted.

major comments (3)
  1. [B.2 (proof of Theorem 3.5)] The proof does not establish the claimed d-dimensional rate. The prior-concentration estimate is correctly quoted with a kappa^d factor, but after restricting kappa to the interval [(C0/epsilon)^{1/alpha}, 2(C0/epsilon)^{1/alpha}], the displayed lower bound becomes exp(-C4 epsilon^{-1/alpha} log^C epsilon) instead of exp(-C4 epsilon^{-d/alpha} log^C epsilon), because kappa^d is of order epsilon^{-d/alpha} on that interval. Similarly, the metric-entropy conclusion is stated as log N(epsilon, F_N^G, ||.||_infty) lesssim epsilon^{-1/alpha}; in the cited bound of [39] the factor is r^d, and with r as epsilon^{-1/alpha} this gives epsilon^{-d/alpha}. The complement-probability bound is also displayed with epsilon^{-1/alpha}. Balancing the written exponent epsilon^{-1/alpha} against n epsilon^2 yields only the rate n^{-alpha/(2alpha+1)}, which is slower than the claimed n^{-alpha/(2alpha+d)} for d>1. As written, the theorem is proved only for d=1; please restore the missing d factors throughout, or restrict Theorem 3.5 to d=1.
  2. [Theorem 3.4 and B.1] The statement of Theorem 3.4 and its proof appear to point in opposite directions. The proof derives a small prior-ratio bound P(||f_N - f*||_2 < xi_n)/P(||f_N - f*||_infty < epsilon_n) lesssim exp(-2 n epsilon_n^2) and concludes that xi_n is a lower bound on the contraction rate. In the Castillo-type lower-bound argument on which the proof is modeled, such a ratio bound implies that the posterior mass in the L2-ball of radius xi_n tends to zero, not to one. As printed, equation (9) states E_{f*} Pi_N(||f_N - f*||_2 < epsilon_n | D_n) -> 1, which is the opposite conclusion. Please correct (9) and the surrounding text so that the theorem's statement matches the lower-bound argument, and specify precisely whether the intended claim is that the posterior cannot contract faster than the displayed rates.
  3. [Section 5] The numerical studies do not satisfy the assumptions of Theorem 3.5, so the claim that they support the theory is currently overstated. The GPI simulations use a Matérn parent kernel (stated in the first paragraph of Section 5), whereas Theorem 3.5 is proved for the squared-exponential kernel (10). In addition, the resolution priors are p(N) proportional to exp(-4N) on {4,8,12,20,30} in Section 5.1 and a discrete uniform prior on {6,8,10,14,18} in Section 5.3; neither satisfies the polynomial-tail condition (12). Please either run simulations within the theorem's scope or explicitly frame the numerics as an exploratory check rather than as a validation of the rate theorem.
minor comments (4)
  1. [Throughout] There are several typographical and notational issues: equation (10) contains '||x|\|_2^2', 'Topelitz' appears for 'Toeplitz' in Section 2.2, and Remark 3.1 contains 'appoach'. The Hölder space notation is also inconsistent in places (C^alpha versus C^{\alpha}).
  2. [B.2, equation (49)] The first sufficient condition is displayed as P(||f_N - f*||_infty < epsilon) greatersim exp{-epsilon^{d/(2alpha+d)} log^C(1/epsilon)}, which has a positive exponent and cannot be the intended lower bound; it should presumably read exp(-epsilon^{-d/(2alpha+d)} log^C(1/epsilon)) or an equivalent negative-exponent form.
  3. [Lemma 3.2] Lemma 3.2 is stated only for d=1, while Theorem 3.5 is for general d. Please state whether a d-dimensional analogue holds, or clarify that the lemma is used only as motivation for the choice N ~ epsilon^{-1/(alpha wedge 2)}.
  4. [Section 4] The complexity statement O(n + N^{3d}) is correct only if N denotes the number of grid points per coordinate; please make this explicit, since the total number of basis functions is (N+1)^d and a reader could otherwise infer a different scaling.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Theorem 3.5's rate rests on external van der Vaart–van Zanten results, not on fitted quantities or author-derived uniqueness claims.

full rationale

The central results are assembled from external benchmarks rather than from the paper's own fitted quantities. Theorem 3.5's KL small-ball bound uses the RBF prior-concentration estimate from van der Vaart and van Zanten (2009), with only the fluctuation control in Lemma A.8/A.9 and the resolution-prior tail condition (12) added by the authors; those additions are genuine analytical steps, not restatements of the desired rate. The entropy and sieve arguments likewise reduce to the external entropy bound for the RBF sieve through the inequality N(eps, F_N^G, ||.||_inf) <= N(eps, G, ||.||_inf), which is a real mapping inequality. The SPDE lower-bound theorem cites [4] (two authors overlap) but only for a proof technique, with the key lower-bound Lemma 1 of [6] being independent external work, so the self-citation is not load-bearing. The computational section cites [27] for fixed-N Toeplitz sampling; Algorithm 1 is an extension and is not used to prove the rates. Section 5.1's support selection from posterior marginal plots is calibration, not a prediction, and it does not feed into Theorem 3.5. The mismatch between the simulation priors (finite support, exponential/uniform) and condition (12) is a scope/external-validity issue, not circularity. The dimensional exponent discrepancy noted in Appendix B.2 (displayed eps^{-1/alpha} bounds where eps^{-d/alpha} is needed for d>1) is a serious proof gap in the multivariate claim, but a proof gap is distinct from circularity: no equation in the derivation is equivalent to the theorem by construction, and no parameter is fitted and then renamed a prediction. The paper therefore contains no circular step, and the appropriate circularity score is 0.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The theoretical results depend on standard Bayesian nonparametric machinery and on two imported concentration bounds, plus a Holder-smoothness domain assumption. The simulation section introduces data-calibrated priors (N support, kappa hyperparameters) that are not part of the theorem. No new particles, forces, or entities are postulated.

free parameters (3)
  • Resolution prior support and hyperparameter = N in {4,8,12,20,30}, p(N) proportional to exp(-4N) (Sec 5.1); N in {6,8,10,14,18}, discrete uniform (Sec 5.3)
    Chosen by inspecting posterior marginal plots in Section 5.1; these are data-calibrated and fall outside theorem condition (12).
  • Bandwidth prior hyperparameters = Gamma(3,1/3) (Sec 5.1); Gamma(5,1/5) (Sec 5.3)
    Chosen from p(kappa | D_n) marginal plots; these satisfy the gamma-type condition (11), so only the N prior is non-compliant.
  • Matern smoothness parameter nu (simulation) = nu = 1.5 or 3.5 for GPI; nu = 7 for GP baseline; beta = 2 for SPDE
    Fixed by user; the GPI theorem assumes an RBF kernel, so using Matern in simulations deviates from the theoretical setup.
assumptions (5)
  • standard math Standard posterior contraction theory for Gaussian process priors (Theorem 3.3 of van der Vaart and van Zanten 2008) is used as the framework to translate prior concentration and entropy bounds into posterior contraction rates.
    Invoked in Section B.2 for Theorem 3.5.
  • standard math The parent squared-exponential GP with inverse-bandwidth kappa and gamma-type prior satisfies the prior concentration lower bound from van der Vaart and van Zanten (2009).
    Used in Section B.2 to lower bound P(||f* - f||_infinity < epsilon/2); this is an external theorem, not re-proved.
  • domain assumption The fluctuation bound for the GP over small grid cells holds with a universal constant K (Lemma A.8, based on Adler's continuity theorem).
    This is a regularity and constant assumption; the proof imports K without explicit value.
  • domain assumption True function lies in Holder class C^alpha[0,1]^d, noise variance sigma is known, and design points are fixed.
    Stated in Section 2 and Theorem 3.5.
  • ad hoc to paper Uniform nonlinearity (UN) condition for the SPDE lower bound (Theorem 3.4).
    Introduced by the authors to relate L2 and sup-norm geometry; only sufficient, and excludes flat or saddle functions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Adaptive Resolution for Finite-Rank Gaussian Processes." pith.science (2026). https://pith.science/paper/VH34NPMO

@misc{pith2026250524066,
  author       = {Pith},
  title        = {Pith review of: Adaptive Resolution for Finite-Rank Gaussian Processes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VH34NPMO}},
  note         = {Machine review of arXiv:2505.24066}
}
abstract

Finite-rank approximations are widely used to scale Gaussian process (GP) regression, but their posterior behavior can differ from that of the corresponding parent GP prior. We study a class of finite-rank GP priors built from locally supported basis expansions with dependent Gaussian coefficients. Our framework covers finite-element approximations based on the stochastic partial differential equation (SPDE) representation of Mat\'ern GPs and regular-grid GP interpolation schemes. We show that, with a suitable prior on the resolution parameter $N$, these finite-rank expansions inherit the same posterior contraction rate as the corresponding parent GP prior under the same bandwidth specification used for that parent prior. Consequently, the interpolation construction under a squared-exponential parent GP attains the minimax-optimal rate up to logarithmic factors under a hierarchical prior on the bandwidth parameter and on $N$, while the SPDE construction attains the same rate under a bandwidth scaling depending on the sample size and the smoothness of the true function, together with a prior on $N$. We also develop a posterior sampler for the hierarchical interpolation model that jointly updates the resolution and bandwidth parameters, and we provide numerical studies that support the theory.

Figures

Figures reproduced from arXiv: 2505.24066 by the authors.

Figure 1
Figure 1. Marginal posterior distributions of p(N | Dn) (left) and p(κ | Dn) (right) for f1 under n ∈ {200, 500, 1000}. samples of f, where f ∼ GP(0, Kκ,ν). 5.1 Adaptive parameter calibration for GPI method Selecting the supports of the parameters N and κ are crucial. For this, we utilize the posterior marginal density plots of p(N | Dn) and p(κ | Dn). We first choose several values of N and κ, and for every possible pair (N,… view at source ↗
Figure 2
Figure 2. Prediction on f1 based on GPI method with different values of ν. The top row corresponds to ν = 1.5, and bottom row to ν = 3.5 Columns show result for n = 200 (left), n = 500 (middle), and n = 1000 (right). Dark gray regions indicate pointwise 95% credible intervals, and the light gray areas indicate 90% intervals. The red lines represent the true function, and the blue lines represent posterior mean. mean squared e… view at source ↗
Figure 3
Figure 3. Prediction on f2 under SPDE method with β = 2 (first row) and Mat´ern GP with ν = 7 (second row). The columns corresponds n = 200 (left), n = 500 (middle), and n = 1000 (right), under fixed κ = 10. Dark gray regions indicate pointwise 95% credible intervals, and the light gray areas indicate 90% intervals. The red line represents the true function , and the blue line represents posterior mean. −9 −8 −7 −6 200 500 10… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: log of AMSE under SPDE method (β = 2) and GP regression (ν = 7) for f2. Results are shown for 6 equidistant κ varying from 5 (top left) to 30 (bottom right). AMSE is computed based on 50 replications of Dn. 16 [PITH_FULL_IMAGE:figures/full_fig_p016_4.png]
Figure 5
Figure 5. Figure 5: log of AMSE under GPI method and the GP regression for [PITH_FULL_IMAGE:figures/full_fig_p017_5.png]
Figure 6
Figure 6. Figure 6: Prediction of f3 under GPI method for n = 500. The heatmaps show the true function f3 (top left), the posterior mean (top right), the lower (bottom left) and upper bound (bottom right) of the pointwise 95% credible interval. −4 −2 0 200 500 1000 2000 n log time(s) meth…
Figure 7
Figure 7. Figure 7: Time comparison between the GPI method and GP regression for obtaining 10 posterior [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 42 canonical work pages

  1. [39]

    Adaptive bayesian estimation using a gaussian random field with inverse gamma bandwidth

    A W van der Vaart and JH van Zanten. Adaptive bayesian estimation using a gaussian random field with inverse gamma bandwidth. The Annals of Statistics , 37(5B):2655–2675, 2009

    work page 2009

  2. [1]

    R.J. Adler. An Introduction to Continuity, Extrema, and Related Topics for General Gaussian Processes. IMS Lecture Notes. Institute of Mathematical Statistics, 1990

    work page 1990

  3. [2]

    Block toeplitz matrix inversion

    Hirotugu Akaike. Block toeplitz matrix inversion. SIAM Journal on Applied Mathematics , 24(2):234–241, 1973

    work page 1973

  4. [3]

    Gaussian predictive process models for large spatial data sets

    Sudipto Banerjee, Alan E Gelfand, Andrew O Finley, and Huiyan Sang. Gaussian predictive process models for large spatial data sets. Journal of the Royal Statistical Society Series B: Statistical Methodology, 70(4):825–848, 2008

    work page 2008

  5. [4]

    Anisotropic function estimation using multi-bandwidth gaussian processes

    Anirban Bhattacharya, Debdeep Pati, and David Dunson. Anisotropic function estimation using multi-bandwidth gaussian processes. Annals of statistics , 42(1):352, 2014

    work page 2014

  6. [5]

    Spatial models generated by nested stochastic partial differen- tial equations, with an application to global ozone mapping

    David Bolin and Finn Lindgren. Spatial models generated by nested stochastic partial differen- tial equations, with an application to global ozone mapping. The Annals of Applied Statistics , pages 523–550, 2011

    work page 2011

  7. [6]

    Lower bounds for posterior rates with gaussian process priors

    Isma¨ el CASTILLO. Lower bounds for posterior rates with gaussian process priors. Electronic Journal of Statistics , 2:1281–1299, 2008

    work page 2008

  8. [7]

    Fixed rank kriging for very large spatial data sets

    Noel Cressie and Gardar Johannesson. Fixed rank kriging for very large spatial data sets. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , 70(1):209–226, 2008

    2008

Show all 46 references
  1. [8]

    Hierarchical nearest- neighbor gaussian process models for large geostatistical datasets

    Abhirup Datta, Sudipto Banerjee, Andrew O Finley, and Alan E Gelfand. Hierarchical nearest- neighbor gaussian process models for large geostatistical datasets. Journal of the American Statistical Association, 111(514):800–812, 2016

    2016

  2. [9]

    de Jonge and J

    R. de Jonge and J. H. van Zanten. Adaptive nonparametric Bayesian inference using location- scale mixture priors. The Annals of Statistics , 38(6):3300 – 3320, 2010

    2010

  3. [10]

    de Jonge and J.H

    R. de Jonge and J.H. van Zanten. Adaptive estimation of multivariate functions using condi- tionally Gaussian tensor-product spline priors. Electronic Journal of Statistics , 6(none):1984 – 2001, 2012

    1984

  4. [11]

    Adaptive estimation of multivariate functions using condi- tionally gaussian tensor-product spline priors

    R de Jonge and JH van Zanten. Adaptive estimation of multivariate functions using condi- tionally gaussian tensor-product spline priors. Electronic Journal of Statistics , 6, 2012. 47

    2012

  5. [12]

    Posterior concentration for gaussian process priors un- der rescaled mat \’ern and confluent hypergeometric covariance functions

    Xiao Fang and Anindya Bhadra. Posterior concentration for gaussian process priors un- der rescaled mat \’ern and confluent hypergeometric covariance functions. arXiv preprint arXiv:2312.07502, 2023

    2023 arXiv

  6. [13]

    Approximate likelihood for large irregularly spaced spatial data

    Montserrat Fuentes. Approximate likelihood for large irregularly spaced spatial data. Journal of the American Statistical Association , 102(477):321–331, 2007

    2007

  7. [14]

    Covariance tapering for interpolation of large spatial datasets

    Reinhard Furrer, Marc G Genton, and Douglas Nychka. Covariance tapering for interpolation of large spatial datasets. Journal of Computational and Graphical Statistics , 15(3):502–523, 2006

    2006

  8. [15]

    Distributed bayesian inference in massive spatial data

    Rajarshi Guhaniyogi, Cheng Li, Terrance Savitsky, and Sanvesh Srivastava. Distributed bayesian inference in massive spatial data. Statistical science, 38(2):262–284, 2023

    2023

  9. [16]

    A case study competition among methods for analyzing large spatial data

    Matthew J Heaton, Abhirup Datta, Andrew O Finley, Reinhard Furrer, Joseph Guinness, Ra- jarshi Guhaniyogi, Florian Gerber, Robert B Gramacy, Dorit Hammerling, Matthias Katzfuss, et al. A case study competition among methods for analyzing large spatial data. Journal of Agricult...

    2019

  10. [17]

    Space and space-time modeling using process convolutions

    Dave Higdon. Space and space-time modeling using process convolutions. In Quantitative methods for current environmental issues , pages 37–56. Springer, 2002

    2002

  11. [18]

    Asymptotic properties of vecchia approximation for gaussian processes

    Myeongjong Kang, Florian Sch¨ afer, Joseph Guinness, and Matthias Katzfuss. Asymptotic properties of vecchia approximation for gaussian processes. arXiv preprint arXiv:2401.15813 , 2024

    2024 arXiv

  12. [19]

    On the asymptotic representation of the euler gamma function by ramanujan

    Ekatherina A Karatsuba. On the asymptotic representation of the euler gamma function by ramanujan. Journal of computational and applied mathematics , 135(2):225–240, 2001

    2001

  13. [20]

    Eigenvalues of tridiagonal pseudo- toeplitz matrices

    Devadatta Kulkarni, Darrell Schmidt, and Sze-Kai Tsui. Eigenvalues of tridiagonal pseudo- toeplitz matrices. Linear Algebra and its Applications , 297:63–80, 1999

    1999

  14. [21]

    Lectures on Gaussian processes

    Mikhail Lifshits and Mikhail Lifshits. Lectures on Gaussian processes. Springer, 2012

    2012

  15. [22]

    An explicit link between gaussian fields and gaussian markov random fields: the stochastic partial differential equation approach

    Finn Lindgren, H ˚ avard Rue, and Johan Lindstr¨ om. An explicit link between gaussian fields and gaussian markov random fields: the stochastic partial differential equation approach. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , 73(4):423–498, 2011

    2011

  16. [23]

    Gaussian process emulators for computer experiments with inequality constraints

    Hassan Maatouk and Xavier Bay. Gaussian process emulators for computer experiments with inequality constraints. Mathematical Geosciences, 49:557–582, 2017

    2017

  17. [24]

    Inequalities for the incomplete gamma function

    Pierpaolo Natalini and Biagio Palumbo. Inequalities for the incomplete gamma function. Math. Inequal. Appl, 3(1):69–77, 2000

    2000

  18. [25]

    A multiresolution gaussian process model for the analysis of large spatial datasets

    Douglas Nychka, Soutir Bandyopadhyay, Dorit Hammerling, Finn Lindgren, and Stephan Sain. A multiresolution gaussian process model for the analysis of large spatial datasets. Journal of Computational and Graphical Statistics , 24(2):579–599, 2015. 48

    2015

  19. [26]

    Gaussian processes for machine learning, volume 1

    Carl Edward Rasmussen, Christopher KI Williams, et al. Gaussian processes for machine learning, volume 1. Springer, 2006

    2006

  20. [27]

    Efficient bayesian shape-restricted function estimation with constrained gaussian process priors

    Pallavi Ray, Debdeep Pati, and Anirban Bhattacharya. Efficient bayesian shape-restricted function estimation with constrained gaussian process priors. Statistics and Computing , 30:839–853, 2020

    2020

  21. [28]

    Rectangular confidence regions for the means of multivariate normal distribu- tions

    Zbynˇ ekˇSid´ ak. Rectangular confidence regions for the means of multivariate normal distribu- tions. Journal of the American statistical association , 62(318):626–633, 1967

    1967

  22. [29]

    Packing hyper- spheres in high-dimensional euclidean spaces

    Monica Skoge, Aleksandar Donev, Frank H Stillinger, and Salvatore Torquato. Packing hyper- spheres in high-dimensional euclidean spaces. Physical Review E—Statistical, Nonlinear, and Soft Matter Physics , 74(4):041127, 2006

    2006

  23. [30]

    Approximating likelihoods for large spatial data sets

    Michael L Stein, Zhiyi Chi, and Leah J Welty. Approximating likelihoods for large spatial data sets. Journal of the Royal Statistical Society Series B: Statistical Methodology , 66(2):275–296, 2004

    2004

  24. [31]

    Geostatistics for large datasets

    Ying Sun, Bo Li, and Marc G Genton. Geostatistics for large datasets. In Advances and challenges in space-time modelling of natural events , pages 55–77. Springer, 2012

    2012

  25. [32]

    An asymptotic analysis of distributed nonparametric methods

    Botond Szab´ o and Harry Van Zanten. An asymptotic analysis of distributed nonparametric methods. Journal of Machine Learning Research , 20(87):1–30, 2019

    2019

  26. [33]

    Adaptive distributed methods under communication constraints

    Botond Szab´ o and Harry van Zanten. Adaptive distributed methods under communication constraints. The Annals of Statistics , 48(4):2347 – 2380, 2020

    2020

  27. [34]

    Vecchia gaussian processes: Probabilistic properties, minimax rates and methodological developments

    Botond Szabo and Yichen Zhu. Vecchia gaussian processes: Probabilistic properties, minimax rates and methodological developments. arXiv preprint arXiv:2410.10649 , 2024

    2024

  28. [35]

    Tsybakov

    Alexandre B. Tsybakov. Introduction to Nonparametric Estimation. Springer Series in Statis- tics. Springer, 2008

    2008

  29. [36]

    A. W. van der Vaart and J. H. van Zanten. Rates of contraction of posterior distributions based on Gaussian process priors. The Annals of Statistics , 36(3):1435 – 1463, 2008

    2008

  30. [37]

    Information rates of nonparametric gaussian process methods

    Aad Van Der Vaart and Harry Van Zanten. Information rates of nonparametric gaussian process methods. Journal of Machine Learning Research , 12(6), 2011

    2011

  31. [38]

    Reproducing kernel hilbert spaces of gaussian priors

    Aad W van der Vaart, J Harry van Zanten, et al. Reproducing kernel hilbert spaces of gaussian priors. IMS Collections , 3:200–222, 2008

    2008

  32. [40]

    An overview of composite likelihood methods

    Cristiano Varin, Nancy Reid, and David Firth. An overview of composite likelihood methods. Statistica Sinica, pages 5–42, 2011. 49

    2011

  33. [41]

    Estimation and model identification for continuous spatial processes

    Aldo V Vecchia. Estimation and model identification for continuous spatial processes. Journal of the Royal Statistical Society Series B: Statistical Methodology , 50(2):297–312, 1988

    1988

  34. [42]

    On stationary processes in the plane

    Peter Whittle. On stationary processes in the plane. Biometrika, pages 434–449, 1954

    1954

  35. [43]

    Reex- amining the proton-radius problem using constrained gaussian processes

    Shuang Zhou, P Giulani, J Piekarewicz, Anirban Bhattacharya, and Debdeep Pati. Reex- amining the proton-radius problem using constrained gaussian processes. Physical Review C , 99(5):055202, 2019

    2019

  36. [44]

    A mass-shifting phe- nomenon of truncated multivariate normal priors

    Shuang Zhou, Pallavi Ray, Debdeep Pati, and Anirban Bhattacharya. A mass-shifting phe- nomenon of truncated multivariate normal priors. Journal of the American Statistical Associ- ation, 119(545):582–596, 2024

    2024

  37. [45]

    Radial neighbours for provably accurate scalable approximations of gaussian processes

    Yichen Zhu, Michele Peruzzi, Cheng Li, and David B Dunson. Radial neighbours for provably accurate scalable approximations of gaussian processes. Biometrika, page asae029, 2024

    2024

  38. [46]

    Trigonometric series, volume 1

    Antoni Zygmund. Trigonometric series, volume 1. Cambridge university press, 2002. 50

    2002

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.