Pith. sign in

REVIEW 3 major objections 6 minor 50 references

Bayesian Inference for Spatially-Temporally Misaligned Data Using Predictive Stacking

T0 review · 3 major / 6 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 Regressing a county-level outcome on a noisy point-referenced exposure can be done exactly, without MCMC, by cutting outcome-to-exposure feedback and stacking analytic posteriors over weakly identified parameters.

desk verdict Useful extension of predictive stacking to spatio-temporal misalignment, with a clean cut-model derivation; needs revision on the data-analysis details and a sensitivity check on the cut. read the letter →

arxiv 2505.24397 v1 pith:O2EUYKNE submitted 2025-05-30 stat.ME stat.AP

Soumyakanti Pan , Sudipto Banerjee This is my paper · ORCID
classification stat.MEstat.AP MSC 62F1562M3062P10
keywords spatial-temporalmisalignmentchangeofsupportpredictivestackingmodularBayesianinferencecutmodelweakidentifiabilityGaussianprocessozoneandasthma
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

Health studies often pair an outcome observed on coarse blocks (a county, a year) with an exposure observed at points and short intervals (ozone monitors, monthly averages). This paper tries to establish a Bayesian way to regress the former on the latter without first imputing exposure and without Markov chain Monte Carlo: model both as views of one latent Gaussian process, cut the feedback from the outcome to the latent exposure, and combine the resulting analytic posteriors by predictive stacking on a grid of poorly identified parameters. The payoff would be a tractable engine for exposure-health associations that propagates every source of uncertainty, from monitoring noise to temporal aggregation to spatial interpolation, into the final coefficient. Applied to California ozone and asthma emergency-department rates, it finds a weak negative ozone association and sizable racial disparities.

What carries the argument

Three objects carry the argument. (1) The multi-resolution change-of-support construction: the latent process is defined at every resolution by stochastic integrals, $Z(\tilde\ell)=|\tilde I|^{-1}\int_{\tilde I}Z(\tilde s,t)\,dt$ for point-month averages and $Z(L)=(|B||I|)^{-1}\int_I\int_B Z(s,t)\,ds\,dt$ for county-year blocks, so the joint law of all resolutions is a single Gaussian distribution with integrated covariance kernels. (2) Assumption 1, the modularization cut, which splits the posterior into a regression module and a spatial-temporal module and is what makes both the analytic posterior and the exposure-only stacking objective possible. (3) Predictive stacking over candidate models $M_g$ fixed at grid values of $(\phi,\delta^2)$: weights $\alpha_g$ maximize the log leave-one-out predictive density, computed exactly via a closed-form $t$-density with rank-one Cholesky updates or approximately via Pareto-smoothed importance sampling, and the stacked posterior $\tilde p(\cdot\mid X,Y)=\sum_g\hat\alpha_g p(\cdot\mid X,Y,M_g)$ carries all downstream inference. A supporting identity, Proposition 1, gives closed-form double integrals of the exponential temporal correlation over intervals, removing all numerical integration over the time dimension.

What would settle it

Take a modest simulated dataset, fit both the cut model (stacked) and the full uncut model (MCMC, run long enough to converge), and compare the posterior of the ozone coefficient $\beta_2$; if the two diverge beyond Monte Carlo error, especially when the outcome is simulated to depend on a component of exposure that the monitors measure noisily, the cut assumption is falsified and the divergence's size measures the discarded information. A cheaper real-data diagnostic is to hold out counties and ask whether adding outcome information improves prediction of held-out monitor readings.

Watch

Extended reading notes

Core claim

The paper's central claim is that a Bayesian hierarchical model in which a block-aggregated outcome $Y(L_k)$ and noisy point-interval exposure measurements $X(\tilde\ell_j)$ depend on a shared latent Gaussian process $Z(\ell)$ becomes fully tractable under one modularization assumption: once $X$ is observed, the outcome $Y$ carries no additional information about the latent process. With that cut in place, and with the spatial-temporal correlation parameters $\phi=(\phi_s,\nu,\phi_t)$ and the noise-to-variance ratio $\delta^2$ held fixed, the posterior factorizes into a linear-regression module and a spatial-temporal module, each built from conjugate Gaussian and inverse-gamma pieces, so posterior samples are drawn by composition sampling rather than iterated sampling. Predictive stacking then averages the candidate-model posteriors, assigning weights by maximizing leave-one-out predictive densities, which are available in closed form as $t$-densities. The result is full Bayesian inference, for the latent ozone surface at any spatial or temporal resolution and for the ozone coefficient in the outcome regression, at a computational cost that stayed under about 90 minutes for 15,725 exposure records.

Load-bearing premise

Everything rests on the cut: once the noisy monthly monitor readings are observed, the asthma outcome is assumed to carry no extra information about true ozone levels, so the exposure surface is learned from monitors alone.

Editorial extensions

If this is right

  • Exposure uncertainty propagates end-to-end: draws of county-year ozone averages feed the outcome regression, so the ozone coefficient's interval includes monitor noise, temporal aggregation, and spatial interpolation instead of treating exposure as known.
  • The stacking step replaces sampling of weakly identified parameters: the posterior is a weighted mixture over grids of $(\phi_s,\nu,\phi_t,\delta^2)$ values, sidestepping the MCMC convergence failures the paper documents for such models.
  • Irregular monitoring schedules need no imputation: the model is fit directly on observed point-month averages, and suppressed or missing outcome records are simply left out of the Gaussian likelihood.
  • In the California analysis the method yields a weakly negative ozone-asthma association (factor 0.97 per 0.005 ppm, 95% interval 0.94 to 1.00), an annual 10% decline in ED visit rates, and a Black-to-White relative rate near 3.6.
  • The stacked posterior predicts the latent ozone surface at arbitrary resolutions, from daily points to quarterly blocks to annual counties, with uncertainty bands; in simulation it modestly beats interpolation-based alternatives on WAIC while uniquely quantifying uncertainty.

Reading between the lines

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

  • A consequence the authors leave implicit is that the cut is the price of tractability: whenever the outcome plausibly reveals exposure beyond the monitors, such as health events that flag unmeasured pollution episodes, the ozone coefficient could be biased, and a sensitivity analysis that relaxes the cut by re-adding part of the feedback would size that bias.
  • The same template of cut, analytic posteriors, and stacking transfers to other exposures such as PM2.5 and NO2, to other outcome supports such as census tracts or hospital catchments, and to non-Gaussian outcomes through a GLM for the outcome module, with the tractability benefit persisting only while that module stays conjugate.
  • A direct test of Assumption 1 would fit, on a dataset small enough for MCMC to converge, both the stacked cut model and the full uncut model; divergence between the two posteriors of the exposure coefficient beyond Monte Carlo error would indict the cut and measure the information discarded.
  • The application's place-based confounding, with higher inland ozone but lower urban emergency-department utilization, suggests the marginally negative ozone coefficient may reflect access-to-care and co-pollutant patterns rather than a protective ozone effect, and the framework's block-level covariates are the natural place to probe that.
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 / 6 minor

Summary. The manuscript proposes a Bayesian hierarchical model for outcomes observed on spatial-temporal blocks (e.g., county-year) and covariates observed at point locations aggregated over time intervals (e.g., monthly ozone at monitoring stations). A latent Gaussian process links the two resolutions, and, conditional on kernel and noise-ratio hyperparameters, the posterior factorizes under a stated cut assumption into an outcome regression module and an exposure module, yielding conjugate posteriors and a composition sampler. The authors then use predictive stacking over a grid of candidate hyperparameters, with weights chosen from leave-one-out predictive densities for the exposure, and they provide a closed-form temporal-integration result (Proposition 1). The method is illustrated with a simulation and with a California asthma/ozone data analysis. The paper also makes code available as an R package.

Significance. If the claims hold, the paper is a useful extension of predictive stacking to spatial-temporal misalignment, offering a scalable alternative to MCMC by combining conjugate posterior updates with model averaging over weakly identified process parameters. The analytic leave-one-out predictive densities and the closed-form temporal integration in Proposition 1 are concrete technical assets, and the authors provide reproducible code. The main inferential claim—that uncertainty is propagated in a fully model-based way—depends critically on the cut assumption in Section 3.3, so the significance of the contribution will depend on how well the paper quantifies the cost of that cut for the outcome regression of interest.

major comments (3)
  1. [Section 3.3, Eq. (7); Section 4.2, Eqs. (15)-(16); Section 5.4, Table 2] The central claim of fully model-based uncertainty propagation rests on Assumption 1, which severs the feedback from Y to Z_L and Z_tilde. In the full generative model (4)-(5), the likelihood p(Y | Z_L, beta, tau^2) generally carries information about Z_L that is not mediated by X; Assumption 1 is a legitimate cut-model choice, but the paper does not quantify the cost of that cut for the key regression coefficient beta2. The stacking weights in (15) are optimized for leave-one-out prediction of X, not for Y, and the simulation in Section 5.4 reports only WAIC on a single dataset, with no evaluation of bias, coverage, or mean squared error of beta2 under a full-model simulation or under plausible violations of Assumption 1. Given the sparse and spatially unbalanced monitors and the roughly 35% informative-missing outcome records described in Section 2, this risk is material. I request a simulation study, or a sensitivity analysis that feeds Y back into the latent process (e.g., a full-model MCMC on a subset), that reports bias and coverage of beta2 and the resulting stacking weights, so the reader can judge the robustness of the reported ozone coefficient.
  2. [Section 6] The data analysis is not reproducible with the stated grid. The text specifies G_phi_s = {0.3, 0.5, 1}, G_nu = {0.5, 1, 1.5}, and G_delta^2 = {1.5, 2}, which is 3 x 3 x 2 = 18 combinations, yet the manuscript states that the analysis stacks on 54 models. The grid for the temporal decay phi_t is never reported. If phi_t is fixed or selected differently, this should be stated explicitly; otherwise the claimed number of candidate models is inconsistent with the specified grids and the reported stacking weights and runtime are not reproducible.
  3. [Section 6, Eq. (20); Section 3.2, Eq. (4)] The outcome model uses error variance inversely proportional to county area, epsilon_ijt ~ N(0, |B_i|^{-1} tau^2). The outcome is an age-adjusted rate per 10,000 residents; for such rates, the sampling variance should scale with the population at risk or the effective denominator, not with geographic area. As specified, small rural counties with large area but small populations receive artificially small variances, and the credible intervals for the race/ethnicity and year effects in Table 2 may be miscalibrated. Please justify this variance choice or examine sensitivity to a population-based D_L.
minor comments (6)
  1. [Section 3.3, Eq. (7)] The displayed factorization has a typo: the first factor is written as p(beta, sigma^2 | Z_L, Y), but the surrounding text and the subsequent derivation require p(beta, tau^2 | Z_L, Y); the second factor should then contain sigma^2. Please correct.
  2. [Section 6, Table 2] The sentence claiming that all regression coefficients except beta14 have 95% credible intervals excluding zero is inconsistent with the table, where the ozone coefficient beta2 has interval (-0.05, 0.00) and therefore includes zero; the text later acknowledges the weak association, so the summary statement should be revised.
  3. [Section 4, Figure 4] The DAG in Figure 4 contains an unexplained node labeled 'dummy' between Y and Z_L; please clarify what this represents or remove it.
  4. [Section 5.4, Table 1] The WAIC differences between the proposed method and the alternatives are less than 0.3 units, and the comparison is based on a single simulated dataset; adding repeated simulations with standard errors would better support the claim that the method is 'slightly better'.
  5. [Section 2 and Section 3.2] The analysis removes roughly 35% of outcome records with strong spatial and demographic missingness patterns and then treats K = 1510 as fixed; as a complete-case analysis this can introduce bias if missingness depends on the outcome or exposure. A brief remark on this limitation would be appropriate.
  6. [Section 6] Please verify the statement that a negative ozone association is 'consistent with patterns reported in previous studies (Zhu et al., 2003)'; the cited study is commonly associated with positive ozone-asthma associations, and the direction of the reported relationship should be checked.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the exposure model, stacking weights, and outcome regression are distinct quantities, with Assumption 1 an explicit cut rather than a hidden equivalence.

full rationale

The paper's derivation chain is self-contained and does not reduce any central claim to its own inputs. The joint model (4)-(5) is a full generative specification, and the posterior (8) follows from Assumption 1, which explicitly states p(ZL, Z_tilde, gamma, sigma2 | Y, X, phi, delta2) = p(ZL, Z_tilde, gamma, sigma2 | X, phi, delta2). This is a deliberate modularization or cut model, not a hidden identity: the outcome Y enters only through the Module 1 posterior p(beta, tau2 | ZL, Y), while ZL and Z_tilde are estimated from X alone. The stacking weights in (15) are optimized for leave-one-out prediction of X, and Assumption 1 is openly invoked to justify computing them solely from Module 2; the subsequent stacked posterior (16) then applies these weights to combine the outcome-relevant posterior distributions. Although this two-stage construction has statistical implications for the interpretation of beta2, it is not circular because the beta2 posterior is not defined in terms of the stacking objective. The grid choices for phi and delta2 are guided by the same exposure data and by the authors' prior stacking papers (e.g., Zhang et al. 2024), but these are practical tuning heuristics rather than load-bearing derivations, and the leave-one-out predictive densities in (17) are derived within the paper. No uniqueness theorem is imported from the authors, and no fitted parameter is renamed as a prediction. The simulation study and WAIC comparison are conducted against external benchmarks (MBA and kriging), providing independent validation. The reader's concern about Assumption 1 is a substantive modeling assumption and potential bias source, but it is explicitly stated and not a circularity. Self-citations are present but not load-bearing; the central methodological contribution—closed-form stacking for spatially-temporally misaligned data—does not reduce to those citations.

Assumptions & free parameters 5 free parameters · 7 assumptions · 0 invented entities

The method rests on a small number of explicit assumptions: the cut in Assumption 1, single-interval temporal aggregation, separable covariance, and a discrete candidate grid. The main free choices are the grid points for phi and delta2; the temporal-decay grid in the data analysis is not fully reported. The outcome error model and complete-case treatment of suppressed rates are assumptions that affect the applied results and should be revisited.

free parameters (5)
  • Candidate grid G_phi_s (spatial decay) = simulation {2,3,5}; data {0.3,0.5,1}
    Chosen by hand so effective ranges span 20-70% of maximum distance (Section 4.1); the stacking estimate depends on these values.
  • Candidate grid G_phi_t (temporal decay) = simulation {0.3,0.5,1}; data values not reported
    Required to form the claimed 54 models in Section 6, but omitted; the exact grid is not reproducible from the text.
  • Candidate grid G_nu (Matern smoothness) = {0.5,1,1.5}
    Customary fixed values for smoothness (Section 4.1).
  • Candidate grid G_delta2 (noise-to-variance ratio) = simulation {0.75,1.5}; data {1.5,2}
    Selected from empirical semivariogram estimates of nugget and partial sill (Section 4.1); anchors the weakly identified noise ratio.
  • Prior hyperparameters (IG scales, V_beta, V_gamma) = a=b=0.1 or 0.01, V_beta=1e3 I
    Weakly informative fixed hyperparameters in Section 4.1 and Section 6; influence the posterior but are not fitted to data.
assumptions (7)
  • domain assumption Assumption 1 (cut): p(Z_L,Z_tilde,gamma,sigma2 | Y,X,phi,delta2) equals p(Z_L,Z_tilde,gamma,sigma2 | X,phi,delta2)
    Section 3.3, Eq. (7); removes feedback from Y to the exposure process and makes analytic posteriors and X-only stacking possible.
  • domain assumption Assumption 2: each temporal block is a single interval (a,b)
    Section 3.5; enables Proposition 1 closed-form temporal integrals.
  • domain assumption Separable covariance C = Matern_spatial * exponential_temporal
    Eq. (12); restricts dependence structure and simplifies computation of integrated covariances.
  • domain assumption Complete-case missingness is ignorable for outcome rates
    Section 2 and Appendix A; about 35% of outcomes are missing, with missingness correlated with race and rurality.
  • domain assumption Outcome error variance decays with inverse spatial-temporal block volume
    Eqs. (4) and (20); rates are population-based, so using county area as a precision proxy is questionable.
  • ad hoc to paper Grid of candidate parameters contains values close to the truth
    Section 4; stacking chooses among fixed candidates and cannot extrapolate beyond the grid.
  • standard math Asymptotic equivalence of KL minimization and LOO log-score maximization
    Section 4.2; justifies optimizing Eq. (15) as a proxy for minimizing KL divergence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Bayesian Inference for Spatially-Temporally Misaligned Data Using Predictive Stacking." pith.science (2026). https://pith.science/paper/O2EUYKNE

@misc{pith2026250524397,
  author       = {Pith},
  title        = {Pith review of: Bayesian Inference for Spatially-Temporally Misaligned Data Using Predictive Stacking},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/O2EUYKNE}},
  note         = {Machine review of arXiv:2505.24397}
}
read the original abstract

Air pollution remains a major environmental risk factor that is often associated with adverse health outcomes. However, quantifying and evaluating its effects on human health is challenging due to the complex nature of exposure data. Recent technological advances have led to the collection of various indicators of air pollution at increasingly high spatial-temporal resolutions (e.g., daily averages of pollutant levels at spatial locations referenced by latitude-longitude). However, health outcomes are typically aggregated over several spatial-temporal coordinates (e.g., annual prevalence for a county) to comply with survey regulations. This article develops a Bayesian hierarchical model to analyze such spatially-temporally misaligned exposure and health outcome data. We introduce Bayesian predictive stacking, which optimally combines multiple predictive spatial-temporal models and avoids iterative estimation algorithms such as Markov chain Monte Carlo that struggle due to convergence issues inflicted by the presence of weakly identified parameters. We apply our proposed method to study the effects of ozone on asthma in the state of California.

Figures

Figures reproduced from arXiv: 2505.24397 by the authors.

Figure 1
Figure 1. Biennial average asthma-related emergency department visit rates (per 10,000) by racial group for each California county from 2015 through 2022. For visualization, rates are averaged over consecutive 2-year periods. 2. Data The data set on adverse health outcomes comprises annual county-level rates of visits to the emergency department (ED) related to asthma per 10,000 residents of California. These data are obtaine… view at source ↗
Figure 2
Figure 2. County boundaries of California and the geographic locations of 200 ozone monitoring sites active from 2015 to 2022. Asian/Pacific Islander and American Indian/Alaskan Native groups [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Interpolated spatial surface of annual (summed over months) ozone con￾centration (in parts per million) for California from 2015 to 2022. The geographic coordinates of the air quality monitoring stations are marked by black circles. and 2022, showing a clear spatial pattern, with the sparsely sampled eastern inland regions record￾ing higher ozone concentrations, demonstrating spatial imbalances in monitoring coverag… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Directed acyclic graph (DAG) illustrating the conditional dependence structure of the hierarchical model (4). Nodes shaded in gray correspond to data that are observed; all other nodes represent unobserved (latent or unknown) quantities. The vertical dashed line denote…
Figure 5
Figure 5. Figure 5: A: Light blue dots represents simulated daily measurements at 100 loca￾tions over a year, whereas dark blue dots denote monthly averages at each site. B: Interpolated spatial surface of a snapshot of the data (denoted by the vertical line in subfigure A) at May 1. C, D…
Figure 6
Figure 6. Figure 6: A: Interpolated spatial surfaces of the quarterly averaged de-trended true spatial-temporal process which simulated the data; B: Median of the stacked posterior predictive distribution ˜p(ZL−µL | X) at the target spatial-temporal blocks. the R package MBA (Finley et al…
Figure 7
Figure 7. Figure 7: 95% credible intervals (in red) of ozone concentration predictions ob￾tained from the stacked posterior ˜p(Z(ℓ) | X), using a periodic Fourier basis mean. 2019 2020 2021 2022 2015 2016 2017 2018 124°W 121°W 118°W 115°W 124°W 121°W 118°W 115°W 124°W 121°W 118°W 115°W 12…
Figure 8
Figure 8. Figure 8: Posterior median of annual ozone concentration predictions at counties of California, obtained from monthly point-referenced observations [PITH_FULL_IMAGE:figures/full_fig_p019_8.png]
Figure 9
Figure 9. Figure 9: A: Posterior distributions of relative effect sizes at constant ozone con￾centration and year; AI/AN: American Indian/ Alaskan native, PI: Pacific Islanders. B: Posterior distributions of the coefficients corresponding to ozone and year. 6.8, 7.9) times the average rat…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 31 canonical work pages

  1. [1]

    Stegun, editors

    Milton Abramowitz and Irene A. Stegun, editors. Handbook of Mathematical Functions with Formulas, Graphs and Mathematical Tables. Dover Publications, Inc., New York, 1965

    work page 1965

  2. [2]

    Carlin, and Alan E

    Sudipto Banerjee, Brad P. Carlin, and Alan E. Gelfand. Hierarchical Modeling and Analysis for Spatial Data. Chapman and Hall/CRC, 2nd edition, 2014. doi:10.1201/b17115

    doi:10.1201/b17115 2014

  3. [3]

    M. J. Bayarri, J. O. Berger, and F. Liu. Modularization in Bayesian analysis, with emphasis on analysis of computer models . Bayesian Analysis, 4 0 (1): 0 119 -- 150, 2009. doi:10.1214/09-BA404

    doi:10.1214/09-ba404 2009

  4. [4]

    Probability and Measure

    Patrick Billingsley. Probability and Measure. John Wiley & Sons, New York, 3rd edition, 1995

    work page 1995

  5. [5]

    Stacked regressions

    Leo Breiman. Stacked regressions. Machine learning, 24 0 (1): 0 49--64, 1996

    work page 1996

  6. [6]

    Asthma emergency department visit rates, 2024

    California Breathing Asthma Program . Asthma emergency department visit rates, 2024. URL https://www.cdph.ca.gov/Programs/CCDPHP/DEODC/EHIB/CPE/Pages/CaliforniaBreathing.aspx. Accessed on 2025-04-18

    work page 2024

  7. [7]

    Bayesian modelling for spatially misaligned health and air pollution data through the INLA-SPDE approach

    Michela Cameletti, Virgilio Gómez-Rubio, and Marta Blangiardo. Bayesian modelling for spatially misaligned health and air pollution data through the INLA-SPDE approach. Spatial Statistics, 31: 0 100353, 2019. doi:10.1016/j.spasta.2019.04.001

    work page doi:10.1016/j.spasta.2019.04.001 2019

  8. [8]

    Clyde and Edwin S

    Merlise A. Clyde and Edwin S. Iversen. B ayesian model averaging in the M-open framework . In Bayesian Theory and Applications , pages 483--498. Oxford University Press, 2013. ISBN 9780199695607. doi:10.1093/acprof:oso/9780199695607.003.0024

    arXiv 2013

Show all 50 references
  1. [9]

    Noel A.C. Cressie. Change of support and the modifiable areal unit problem. Geographical Systems, 3: 0 159--180, 1996

    1996

  2. [10]

    MBA : M ultilevel B -Spline Approximation , 2024

    Andrew Finley, Sudipto Banerjee, and Øyvind Hjelle. MBA : M ultilevel B -Spline Approximation , 2024. URL https://CRAN.R-project.org/package=MBA. R package version 0.1-2

    2024

  3. [11]

    Finley, Sudipto Banerjee, and Alan E

    Andrew O. Finley, Sudipto Banerjee, and Alan E. Gelfand. spbayes for large univariate and multivariate point-referenced spatio-temporal data models. Journal of Statistical Software, 63 0 (13): 0 1–28, 2015. doi:10.18637/jss.v063.i13

    2015 doi

  4. [12]

    Model Determination using Predictive Distributions with Implementation via Sampling-Based Methods

    A E Gelfand, D K Dey, and H Chang. Model Determination using Predictive Distributions with Implementation via Sampling-Based Methods . In Bayesian Statistics 4: Proceedings of the Fourth Valencia International Meeting, Dedicated to the memory of Morris H. DeGroot, 1931–1989 . ...

    1931

  5. [13]

    Gelfand, Li Zhu, and Bradley P

    Alan E. Gelfand, Li Zhu, and Bradley P. Carlin. On the change of support problem for spatio-temporal data. Biostatistics, 2 0 (1): 0 31--45, 2001. ISSN 1465-4644. doi:10.1093/biostatistics/2.1.31

    2001 doi

  6. [14]

    Understanding predictive information criteria for B ayesian models

    Andrew Gelman, Jessica Hwang, and Aki Vehtari. Understanding predictive information criteria for B ayesian models. Statistics and Computing, 24: 0 997--1016, 2013. doi:10.1007/s11222-013-9416-2

    2013 doi

  7. [15]

    Entwistle, Sandie Ha, Mariaelena Gonzalez, Paul Brown, Donald Schweizer, and Ricardo Cisneros

    Hamed Gharibi, Marcela R. Entwistle, Sandie Ha, Mariaelena Gonzalez, Paul Brown, Donald Schweizer, and Ricardo Cisneros. Ozone pollution and asthma emergency department visits in the central valley, california, usa, during june to september of 2015: a time-stratified case-cros...

    2015 arXiv

  8. [16]

    Gneiting and P

    T. Gneiting and P. Guttorp. Continuous-parameter spatio-temporal processes. In A.E. Gelfand, P.J. Diggle, M. Fuentes, and P Guttorp, editors, Handbook of Spatial Statistics, Chapman & Hall CRC Handbooks of Modern Statistical Methods, pages 427--436. Taylor and Francis, 2010

    2010

  9. [17]

    Tilmann Gneiting and Adrian E. Raftery. Strictly proper scoring rules, prediction, and estimation. Journal of the American Statistical Association, 102 0 (477): 0 359--378, 2007. doi:10.1198/016214506000001437

    2007 doi

  10. [18]

    Combining incompatible spatial data

    Carol A Gotway and Linda J Young and. Combining incompatible spatial data. Journal of the American Statistical Association, 97 0 (458): 0 632--648, 2002. doi:10.1198/016214502760047140

    2002 doi

  11. [19]

    Hoeting, David Madigan, Adrian E

    Jennifer A. Hoeting, David Madigan, Adrian E. Raftery, and Chris T. Volinsky. B ayesian model averaging: a tutorial (with comments by M. Clyde, David Draper and E. I. George, and a rejoinder by the authors . Statistical Science, 14 0 (4): 0 382 -- 417, 1999. doi:10.1214/ss/1009212519

    1999

  12. [20]

    Jacob, Lawrence M

    Pierre E. Jacob, Lawrence M. Murray, Chris C. Holmes, and Christian P. Robert. Better together? statistical learning in models made of modules, 2017. URL https://doi.org/10.48550/arXiv.1708.08719

    2017 doi arXiv

  13. [21]

    Fast algorithm for cross-validation of the best linear unbiased predictor

    Tae Yoon Kim, Jeong Soo Park, and Dennis D Cox and. Fast algorithm for cross-validation of the best linear unbiased predictor. Journal of Computational and Graphical Statistics, 11 0 (4): 0 823--835, 2002. doi:10.1198/106186002826

    2002 doi

  14. [22]

    A more efficient rank-one covariance matrix update for evolution strategies

    Oswin Krause and Christian Igel. A more efficient rank-one covariance matrix update for evolution strategies. In Proceedings of the 2015 ACM Conference on Foundations of Genetic Algorithms XIII, FOGA '15, page 129–136, New York, NY, USA, 2015. Association for Computing Machine...

    2015

  15. [23]

    A Bayes Interpretation of Stacking for M -Complete and M -Open Settings

    Tri Le and Bertrand Clarke. A Bayes Interpretation of Stacking for M -Complete and M -Open Settings . Bayesian Analysis, 12 0 (3): 0 807 -- 829, 2017. doi:10.1214/16-BA1023

    2017 doi

  16. [24]

    B ayesian model averaging

    David Madigan, Adrian E Raftery, C Volinsky, and Jennifer Hoeting. B ayesian model averaging. In Proceedings of the AAAI Workshop on Integrating Multiple Learned Models, Portland, OR, pages 77--83, 1996

    1996

  17. [25]

    K. V. Mardia and C. R. Goodall. Spatial-temporal analysis of multivariate environmental monitoring data. Multivariate Environmental Statistics, 1993

    1993

  18. [26]

    Mugglin, Bradley P

    Andrew S. Mugglin, Bradley P. Carlin, and Alan E. Gelfand. Fully model-based approaches for spatially misaligned data. Journal of the American Statistical Association, 95 0 (451): 0 877--887, 2000. doi:10.1080/01621459.2000.10474279

  19. [27]

    Nolte, Patrick Dolwick, Tanya L

    Nicholas Nassikas, Keith Spangler, Neal Fann, Christopher G. Nolte, Patrick Dolwick, Tanya L. Spero, Perry Sheffield, and Gregory A. Wellenius. Ozone-related asthma emergency department visits in the us in a warming climate. Environmental Research, 183: 0 109206, 2020. ISSN 00...

    2020

  20. [28]

    spStack : B ayesian Geostatistics Using Predictive Stacking , 2024

    Soumyakanti Pan and Sudipto Banerjee. spStack : B ayesian Geostatistics Using Predictive Stacking , 2024. URL https://CRAN.R-project.org/package=spStack. R package version 1.0.1

    2024

  21. [29]

    Bradley, and Sudipto Banerjee

    Soumyakanti Pan, Lu Zhang, Jonathan R. Bradley, and Sudipto Banerjee. Bayesian inference for spatial-temporal non-gaussian data using predictive stacking, 2025. URL https://doi.org/10.48550/arXiv.2406.04655

    2025 doi

  22. [30]

    Cuts in bayesian graphical models

    Martyn Plummer. Cuts in bayesian graphical models. Statistics and Computing, 25: 0 37--43, 2014. doi:10.1007/s11222-014-9503-z

    2014 doi

  23. [31]

    B ayesian transfer learning for artificially intelligent geospatial systems: A predictive stacking approach, 2025

    Luca Presicce and Sudipto Banerjee. B ayesian transfer learning for artificially intelligent geospatial systems: A predictive stacking approach, 2025. URL https://doi.org/10.48550/arXiv.2410.09504

    2025 doi arXiv

  24. [32]

    Harrison Quick, Sudipto Banerjee, and Bradley P. Carlin. Modeling temporal gradients in regionally aggregated California asthma hospitalization data . The Annals of Applied Statistics, 7 0 (1): 0 154 -- 176, 2013. doi:10.1214/12-AOAS600

    2013 doi

  25. [33]

    Carlin, and Sudipto Banerjee

    Harrison Quick, Bradley P. Carlin, and Sudipto Banerjee. Heteroscedastic conditional auto-regression models for areally referenced temporal processes for analysing california asthma hospitalization data. Journal of the Royal Statistical Society Series C: Applied Statistics, 64...

    2015 doi

  26. [34]

    geoR : A nalysis of Geostatistical Data , 2024

    Paulo Justiniano Ribeiro Jr , Peter Diggle, Ole Christensen, Martin Schlather, Roger Bivand, and Brian Ripley. geoR : A nalysis of Geostatistical Data , 2024. URL https://CRAN.R-project.org/package=geoR. R package version 1.9-4

    2024

  27. [35]

    Approximate B ayesian Inference for Latent Gaussian models by using Integrated Nested Laplace Approximations

    Håvard Rue, Sara Martino, and Nicolas Chopin. Approximate B ayesian Inference for Latent Gaussian models by using Integrated Nested Laplace Approximations . Journal of the Royal Statistical Society Series B: Statistical Methodology, 71 0 (2): 0 319--392, 2009. ISSN 1369-7412. ...

    2009

  28. [36]

    On identifiability and consistency of the nugget in G aussian spatial process models

    Wenpin Tang, Lu Zhang, and Sudipto Banerjee. On identifiability and consistency of the nugget in G aussian spatial process models. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 83 0 (5): 0 1044--1070, 2021. doi:https://doi.org/10.1111/rssb.12472...

    2021 doi

  29. [37]

    deldir : Delaunay Triangulation and D irichlet ( V oronoi) Tessellation , 2024

    Rolf Turner. deldir : Delaunay Triangulation and D irichlet ( V oronoi) Tessellation , 2024. URL https://CRAN.R-project.org/package=deldir. R package version 2.0-4

    2024

  30. [38]

    Practical B ayesian model evaluation using leave-one-out cross-validation and waic

    Aki Vehtari, Andrew Gelman, and Jonah Gabry. Practical B ayesian model evaluation using leave-one-out cross-validation and waic. Statistics and Computing, 27 0 (5): 0 1413–1432, 2017. ISSN 0960-3174. doi:10.1007/s11222-016-9696-4

    2017 doi

  31. [39]

    loo: Efficient leave-one-out cross-validation and waic for bayesian models, 2024 a

    Aki Vehtari, Jonah Gabry, Måns Magnusson, Yuling Yao, Paul-Christian Bürkner, Topi Paananen, and Andrew Gelman. loo: Efficient leave-one-out cross-validation and waic for bayesian models, 2024 a . URL https://mc-stan.org/loo/. R package version 2.8.0

    2024

  32. [40]

    Pareto smoothed importance sampling

    Aki Vehtari, Daniel Simpson, Andrew Gelman, Yuling Yao, and Jonah Gabry. Pareto smoothed importance sampling. Journal of Machine Learning Research, 25 0 (72): 0 1--58, 2024 b . URL http://jmlr.org/papers/v25/19-556.html

    2024

  33. [41]

    Process-based inference for spatial energetics using B ayesian predictive stacking, 2024

    Tomoya Wakayama and Sudipto Banerjee. Process-based inference for spatial energetics using B ayesian predictive stacking, 2024. URL https://doi.org/10.48550/arXiv.2405.09906

    2024 doi arXiv

  34. [42]

    Asymptotic equivalence of B ayes cross validation and widely applicable information criterion in singular learning theory

    Sumio Watanabe. Asymptotic equivalence of B ayes cross validation and widely applicable information criterion in singular learning theory. Journal of Machine Learning Research, 11 0 (116): 0 3571--3594, 2010. URL http://jmlr.org/papers/v11/watanabe10a.html

    2010

  35. [43]

    Stacked generalization

    David H Wolpert. Stacked generalization. Neural networks, 5 0 (2): 0 241--259, 1992

    1992

  36. [44]

    Using stacking to average B ayesian predictive distributions (with discussion)

    Yuling Yao, Aki Vehtari, Daniel Simpson, and Andrew Gelman. Using stacking to average B ayesian predictive distributions (with discussion). B ayesian Analysis , 13 0 (3): 0 917--1007, 2018

    2018

  37. [45]

    Inconsistent estimation and asymptotically equal interpolations in model-based geostatistics

    Hao Zhang. Inconsistent estimation and asymptotically equal interpolations in model-based geostatistics. Journal of the American Statistical Association, 99 0 (465): 0 250--261, 2004. doi:10.1198/016214504000000241

    2004 doi

  38. [46]

    Zimmerman

    Hao Zhang and Dale L. Zimmerman. Towards reconciling two asymptotic frameworks in spatial statistics. Biometrika, 92 0 (4): 0 921--936, 2005. ISSN 0006-3444. doi:10.1093/biomet/92.4.921

    2005 doi

  39. [47]

    Stephens

    Jin Zhang and Michael A. Stephens. A new and efficient estimation method for the generalized pareto distribution. Technometrics, 51 0 (3): 0 316--325, 2009. doi:10.1198/tech.2009.08017

    2009 arXiv

  40. [48]

    Bayesian geostatistics using predictive stacking, 2024

    Lu Zhang, Wenpin Tang, and Sudipto Banerjee. Bayesian geostatistics using predictive stacking, 2024. URL https://doi.org/10.48550/arXiv.2304.12414

    2024 doi

  41. [49]

    B ayesian hierarchical models for the combination of spatially misaligned data: A comparison of melding and downscaler approaches using INLA and SPDE

    Ruiman Zhong and Paula Moraga. B ayesian hierarchical models for the combination of spatially misaligned data: A comparison of melding and downscaler approaches using INLA and SPDE . Journal of Agricultural, Biological and Environmental Statistics, 29: 0 110--129, 2023. doi:10...

    2023 doi

  42. [50]

    Carlin, and Alan E

    Li Zhu, Bradley P. Carlin, and Alan E. Gelfand. Hierarchical regression with misaligned spatial data: relating ambient ozone and pediatric asthma er visits in atlanta. Environmetrics, 14 0 (5): 0 537--557, 2003. doi:https://doi.org/10.1002/env.614

    2003 doi

Pith tools

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