Pith. sign in

REVIEW 3 major objections 6 minor 80 references

Bayesian Stacking via Proper Scoring Rule Optimization using a Gibbs Posterior

T0 review · 3 major / 6 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper introduces the stacked Gibbs posterior, a Bayesian distribution over linear-pool weights that optimizes a proper scoring rule, and reports it often beats model averaging and equal weighting in simulations and in the 2023-24 FluSig

desk verdict New and useful Gibbs-posterior stacking method for linear-pool weights; FluSight evidence rests on an unvalidated quantile reconstruction. read the letter →

arxiv 2509.04203 v1 pith:GVS73Y7M submitted 2025-09-04 stat.ME

Spencer Wadsworth , Jarad Niemi This is my paper · ORCID
classification stat.ME MSC 62F1562F1262M20
keywords stackedGibbsposterioroptimallinearpoolingprobabilisticforecastingproperscoringrulescontinuousrankedprobabilityscoreensembleBayesianmodelaveragingforecastcombination
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper's central claim is that a Gibbs posterior over linear-pool weights—probability mass proportional to exp(−η n times the average CRPS) times a Dirichlet prior—is a practical way to do stacking for probabilistic forecasts. Unlike point-optimized stacking or model-averaging posteriors, this approach returns a full distribution over weights, so each component's influence carries uncertainty, and the prior can pull the ensemble toward equal weighting. The authors report that in three simulation studies the stacked Gibbs posterior (SGP) frequently beats Bayesian model averaging, adaptive variable selection, and equal weighting on proper-score and calibration measures, and in the 2023-24 FluSight flu-hospitalization forecasts it ranked first in more regions and weeks than the other methods. The appeal, if the claim holds, is that forecast hubs get a principled way to optimize the exact scoring rule they are judged by without losing the safety of equal-weight pooling.

What carries the argument

The central object is the stacked Gibbs posterior (SGP): a distribution over the weight simplex, π_n(η)(ω) ∝ exp{−η n S_n(ω)} π(ω), where S_n is the average CRPS of the mixture forecast over past outcomes (optionally discounted) and π(ω) is a Dirichlet prior. The prior's full-simplex support satisfies the consistency condition and lets the user regularize toward equal weights by increasing concentration. The load-bearing identity is Proposition 1, CRPS(P̄, y) = Σ w_c E|X_c−y| − ½ ΣΣ w_c w_d E|X_c−X_d|, which turns CRPS evaluation of a mixture into expected absolute differences between components and observations, estimable from draws. The Gibbs-posterior formulation converts weight selection

What would settle it

Take the FluSight data and re-estimate the ensemble weights twice: once using the quantile Gaussian process draws (as in the paper) and once computing CRPS directly from a piecewise-linear CDF on the submitted quantile grid, then compare the SGP, AVS, BMA, and EQW rankings. If the SGP no longer ranks first in 31 of 53 regions, the reported advantage depends on the reconstruction. Separately, in the i.i.d. simulation, check whether the SGP posterior mean tracks the exact CRPS-minimizing weights as n grows; if not, the consistency result is not operative.

Watch

Extended reading notes

Core claim

The core discovery is the stacked Gibbs posterior (SGP), defined as π_n(η)(ω) ∝ exp{−η n S_n(ω)} π(ω), with S_n the empirical CRPS of the linear-pool forecast (optionally discounted for dynamic settings) and π(ω) a Dirichlet prior. Theorem 1 shows that for i.i.d. data and fixed component models the posterior concentrates on the weight vector minimizing expected CRPS. Proposition 1 extends the known normal-mixture CRPS formula to arbitrary continuous mixture components: CRPS(P̄, y) = Σ_c w_c E|X_c−y| − ½ Σ_c Σ_d w_c w_d E|X_c−X_d|, which is what makes the method usable when components are Monte Carlo draws. In the paper's analysis, SGP ensembles score well on CRPS/LogS and PIT uniformity; in

Load-bearing premise

The FluSight result, the primary real-data evidence, assumes the quantile Gaussian process matching model converts submitted quantile forecasts into continuous distributions accurately enough that Monte Carlo CRPS values computed from its posterior draws faithfully represent the original submissions; if that conversion is inaccurate, the reported weight estimates and CRPS rankings are not measuring what they claim.

Editorial extensions

If this is right

  • Forecast hubs gain a way to publish posterior intervals for ensemble weights, making the combination method auditable rather than a single point estimate.
  • A stronger Dirichlet prior centered at equal weights becomes a built-in shrinkage knob; in the SIR simulation the regularized SGP50 variant produced the lowest LogS and CRPS among the compared methods.
  • Proposition 1 lets SGP accept component forecasts as samples, not only closed-form CDFs, so hubs that receive quantiles or posterior draws can still build CRPS-optimal linear pools.
  • Theorem 1 gives static settings a guarantee: with i.i.d. data and fixed models, the posterior provably concentrates on the CRPS-minimizing weights as n grows, so the method does not rely on ad hoc asymptotics in that regime.
  • In the 2023-24 FluSight reevaluation, SGP placed first by mean CRPS in 31 of 53 regions and 14 of 29 weeks, consistent with the claim that it outperforms the equal-weight default often enough to matter in practice.

Reading between the lines

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

  • A same-family extension is to replace CRPS with the weighted interval score (WIS) that FluSight actually uses; because WIS approaches CRPS as the interval set grows, an SGP-like posterior over WIS would directly optimize the competition's headline metric.
  • The fixed Dirichlet prior makes the posterior weight distribution static; a state-space prior on ω would let SGP track regime changes, a direction the paper itself notes is not covered by its i.i.d. theory.
  • The FluSight analysis keeps only teams with complete weekly submissions, so SGP weights are estimated on a survivor set; extending the Gibbs posterior to missing forecasts is a natural way to recover value from dropped teams.
  • Treating the Dirichlet concentration (1 vs 50 in the paper's SGP50) as a tuned shrinkage parameter selected by rolling-origin cross-validation is a natural extension consistent with the paper's learning-rate tuning philosophy.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes a 'stacked Gibbs posterior' (SGP) for combining probabilistic forecasts in a linear pool. The SGP is defined in Eq. (4) as a Gibbs posterior over the simplex of combination weights, with empirical risk based on the CRPS and a prior distribution on the weights. The authors give a proposition expressing the CRPS of a mixture of continuous forecast distributions as a quadratic form in the weights, a consistency theorem for i.i.d. data and fixed component models, three simulation studies (i.i.d. mixture, dynamic mixture, and SIR epidemic forecasts), and an application to the 2023-24 CDC FluSight hospitalization forecasts, comparing SGP with adaptive variable selection (AVS), Bayesian model averaging (BMA), and equal weighting (EQW). The abstract claims that SGP 'often outperforms' the other methods in both simulations and the FluSight analysis.

Significance. If the empirical claims hold, the SGP is a useful addition to the forecast combination toolkit: it gives a principled Bayesian-style posterior over ensemble weights, permits regularization toward equal weighting through the prior, and provides uncertainty quantification for the weights. The consistency theorem and the continuous-mixture CRPS formula (Prop. 1) are useful, and the paper is honest about the scope of the theory, explicitly limiting Theorem 1 to the i.i.d., fixed-model setting. The main value is the clear construction and the Monte Carlo CRPS evaluation for mixtures. However, the central real-data evidence is currently weakened by the unvalidated quantile-to-distribution reconstruction used in the FluSight analysis, and some simulation claims are stronger than the plots support. The authors do not provide code or data, and they do not compare against the most closely related stacking method (Yao et al., 2018).

major comments (3)
  1. [Section 5, Fig. 7, Table 1] The SGP weights and all CRPS scores in the FluSight analysis are computed from 50,000 posterior predictive draws of the quantile Gaussian process (qGP) matching model, not directly from the quantile forecasts submitted by the teams. The manuscript provides no validation that the qGP predictive distribution faithfully represents each team's submitted quantile forecast. If the qGP is misspecified, over- or underdispersed, or dominated by prior uncertainty, then the CRPS rankings in Table 1 and the SGP weights measure the reconstructed distributions, not the actual FluSight forecasts. Because this analysis is the primary real-data evidence for the abstract's outperformance claim, please add a fidelity check (e.g., PIT of the qGP draws against the original quantiles, or a comparison of CRPS computed from qGP draws with WIS computed directly from the 23 submitted quantiles) or substantially s
  2. [Section 8.2, proof of Theorem 1] The proof of Theorem 1 contains a garbled line: 'Now Ω\B_δ ⊂ Ω\B_δ′, so G(Ω\B_δ′)⊂ G(Ω\B_δ′)' — the set on the left should be G(Ω\B_δ). The correction is straightforward: Ω\B_δ ⊂ Ω\B_δ′ implies G(Ω\B_δ) ⊂ G(Ω\B_δ′), hence inf_{Ω\B_δ′} G ≤ inf_{Ω\B_δ} G; since ω* ∉ Ω\B_δ′ and ω* is the unique minimizer, G(ω*) < inf_{Ω\B_δ′} G, giving (20). In addition, the theorem statement should include a finite first-moment condition on the component models (E|X_c| < ∞ for each c), because the domination bound in the proof uses m(y) = E|X| + |y|; the present statement only assumes E|Y| < ∞. Without this, the hypotheses do not guarantee that the CRPS of the mixture is finite.
  3. [Section 4.3 and Section 4 summary] The statement that 'The SGP largely outperformed AVS, BMA, and EQW ensembles' is not supported by the SIR simulation for the default SGP. Figure 6 shows that the plain SGP does not perform as well as the competitors, and the competitive variant SGP50 requires a specially strengthened Dirichlet prior (the text appears to say the prior parameter vector is '504', presumably 50/4) that heavily regularizes toward equal weights. The paper should separate the two claims: a default SGP is not uniformly better; a strongly regularized SGP is. This matters because the FluSight analysis uses the default uninformative Dirichlet prior, so the real-data outperformance of the default method is not explained by this simulation.
minor comments (6)
  1. [Throughout] Editorial notes remain in the text (e.g., 'Jarad: Include lower is better' near Fig. 4), and there are typos such as 'foreacsts' in the abstract, 'Similary' in §8.1, 'peprformed' near Fig. 7, and 'a alternative' in §3.1.
  2. [Section 4.1] The text says 'the empirical risk function from (5) is used'; this should be Eq. (9). Eq. (5) defines the risk minimizer, not the empirical risk.
  3. [Section 4.3] The prior parameter values are written as '14' and '504'; these presumably mean 1/4 and 50/4. Please write them unambiguously, e.g., Dir(1/4, ..., 1/4) and Dir(50/4, ..., 50/4).
  4. [Section 5, Fig. 7] The text states that the left panel shows 90% credible intervals for Rhode Island, while the caption says the figure is for the national level; please clarify which is displayed. Also, define the number of included teams/models in the text (the caption mentions 10) and state the number of regions explicitly.
  5. [Section 5] The Monte Carlo mean formula in the text should be \bar w_{c,t} = M^{-1} \sum_{m=1}^M w^{(m)}_{c,t}; as written it is missing the summation over posterior draws.
  6. [Empirical comparisons] No comparison is made to the original stacking method of Yao et al. (2018) or to a point-estimate CRPS weight optimizer, despite the paper's emphasis on stacking. Adding such a baseline would help calibrate the 'often outperform' claim and is important for positioning the contribution.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the SGP derivation is self-contained and the FluSight qGP preprocessing is an external assumption, not a circular reduction.

full rationale

The SGP is defined directly as a Gibbs posterior (Eq. 4) with a CRPS-based empirical risk; its components (prior, risk, learning rate) are not defined in terms of later results. Proposition 1 (Eq. 15) is a linearity-based corollary of the Gneiting–Raftery CRPS representation (Eq. 14), not a reverse-engineered output. Theorem 1 is a specialization of the Martin–Syring Gibbs posterior consistency theorem; the proof checks the standard uniform-LLN and identifiability conditions rather than importing the conclusion. Simulation comparisons evaluate holdout draws or sequential one-step-ahead forecasts, so the CRPS used for weight selection is not the same data used for evaluation. The FluSight analysis relies on the authors' quantile Gaussian process model (Wadsworth and Niemi 2025b) to convert discrete quantiles to continuous draws, and all CRPS values are Monte Carlo estimates from those draws; this is a potentially fragile preprocessing assumption, but it is not a reduction of the SGP's own derivation to its inputs, and the temporal out-of-sample construction prevents the fitted-input-called-prediction pattern. No load-bearing uniqueness claim or ansatz is imported solely through self-citation. The paper explicitly lists limitations (i.i.d.-only consistency, missing forecasts), further indicating that no hidden circular reliance is being claimed. The use of the same proper scoring rule for fitting and evaluation is standard optimal linear pooling and does not by itself constitute circularity.

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

No new physical or model entities are postulated; the SGP is a distribution over existing weight parameters. The free parameters are the learning rate eta, the discount factor, and the prior concentration, all of which can materially change the reported performance.

free parameters (4)
  • Learning rate eta for SGP/AVS = 15 in section 4.1; 1 in sections 4.3 and 5
    Controls the influence of the empirical CRPS risk. For SGP in section 4.1 eta was fixed at 15 after inspecting how the CRPS changed with eta; in other settings eta=1 was chosen because performance was insensitive. This is a data-informed tuning choice, not derived from theory.
  • Discount factor alpha in dynamic risk (10) = 0.98
    Copied from Lavine et al. (2021); lessens influence of older observations. Not tuned here.
  • Dirichlet prior concentration for SGP50 = 50/4 per component (12.5)
    In section 4.3, a strong prior was chosen to regularize weights toward equality; this is the version that outperformed in the SIR study. The choice is ad hoc.
  • Uninformative Dirichlet prior parameter for default SGP = 1 (vector of ones)
    Standard default; not a fitted value but a modeling choice that affects results.
assumptions (5)
  • standard math Gibbs posterior consistency conditions of Martin and Syring (2022)
    Used in section 3.3 to state and prove Theorem 1.
  • standard math CRPS representation CRPS(P,y)=E|X-y|-1/2 E|X-X'| from Gneiting and Raftery (2007)
    Basis of Proposition 1 and the empirical risk.
  • standard math Lemma 2.4 of Newey and McFadden (1994) for uniform laws of large numbers
    Used in the proof of Theorem 1 to show uniform convergence of the empirical risk.
  • domain assumption The quantile Gaussian process matching model (Wadsworth and Niemi 2025b) faithfully converts FluSight quantile forecasts to continuous distributions
    Section 5 relies on this self-cited model to compute CRPS values; if it is miscalibrated, the SGP weights and reported scores are affected.
  • domain assumption Component forecasts are fixed and data are i.i.d. in Theorem 1
    Theory only covers this setting; the paper acknowledges dynamic scenarios are not covered by the consistency result.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Bayesian Stacking via Proper Scoring Rule Optimization using a Gibbs Posterior." pith.science (2026). https://pith.science/paper/GVS73Y7M

@misc{pith2026250904203,
  author       = {Pith},
  title        = {Pith review of: Bayesian Stacking via Proper Scoring Rule Optimization using a Gibbs Posterior},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GVS73Y7M}},
  note         = {Machine review of arXiv:2509.04203}
}
read the original abstract

In collaborative forecast projects, the combining of multiple probabilistic forecasts into an ensemble is standard practice, with linear pooling being a common combination method. The weighting scheme of a linear pool should be tailored to the specific research question, and weight selection is often performed via optimizing a proper scoring rule. This is known as optimal linear pooling. Besides optimal linear pooling, Bayesian predictive synthesis has emerged as a model probability updating scheme which is more flexible than standard Bayesian model averaging and which provides a Bayesian solution to selecting model weights for a linear pool. In many problems, equally weighted linear pool forecasts often outperform forecasts constructed using sophisticated weight selection methods. Thus regularization to an equal weighting of forecasts may be a valuable addition to any weight selection method. In this manuscript, we introduce an optimal linear pool based on a Gibbs posterior over stacked model weights optimized over a proper scoring rule. The Gibbs posterior extends stacking into a Bayesian framework by allowing for optimal weight solutions to be influenced by a prior distribution, and it also provides uncertainty quantification of weights in the form of a probability distribution. We compare ensemble forecast performance with model averaging methods and equal weighted models in simulation studies and in a real data example from the 2023-24 US Centers for Disease Control FluSight competition. In both the simulation studies and the FluSight analysis, the stacked Gibbs posterior produces ensemble forecasts which often outperform the ensembles of other methods.

Figures

Figures reproduced from arXiv: 2509.04203 by the authors.

Figure 1
Figure 1. United States flu hospitalization forecasts from multiple competing forecast teams (left) for 1-4 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Mixture distribution density function from which data are simulated (black) and 6 candidate [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Examples of estimated densities after weighting of competing models faceted by weighting method [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Boxplots of the mean of 1,000 Monte Carlo LogSs (left) and CRPSs (right) for 500 replicates and [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Plots showing results for one step ahead forecasts for dynamically weighted data for the four [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Plots showing results for one step ahead forecasts for SIR data for the five weighting methods SGP, [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: 90% credible intervals for forecast weights estimated via SGP for the 2023-24 CDC FluSite targetting [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]
Figure 8
Figure 8. Figure 8: An example of CRPS values for different values of [PITH_FULL_IMAGE:figures/full_fig_p021_8.png]
Figure 9
Figure 9. Figure 9: An example of simulated component weights for 100 time steps. The figure shows the simulated [PITH_FULL_IMAGE:figures/full_fig_p021_9.png]
Figure 10
Figure 10. Figure 10: Plots showing results for one step ahead forecasts for SIR data for the four weighting methods [PITH_FULL_IMAGE:figures/full_fig_p022_10.png]
Figure 11
Figure 11. Figure 11: Mean over regions WIS for the four ensemble methods divided by mean WIS for the SGP [PITH_FULL_IMAGE:figures/full_fig_p022_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

80 extracted references · 77 canonical work pages

  1. [1]

    J., Brauer, F., Van den Driessche, P., and Wu, J

    Allen, L. J., Brauer, F., Van den Driessche, P., and Wu, J. (2008). Mathematical Epidemiology , volume 1945. Springer

    work page 2008

  2. [2]

    Bassetti, F., Casarin, R., and Ravazzolo, F. (2018). Bayesian nonparametric calibration and combination of predictive distributions. Journal of the American Statistical Association , 113(522):675--685

    work page 2018

  3. [3]

    Bernardo, J. M. and Smith, A. F. (1994). Bayesian Theory . John Wiley & Sons

    work page 1994

  4. [4]

    and Ziel, F

    Berrisch, J. and Ziel, F. (2023). CRPS learning . Journal of Econometrics , 237(2):105221

    work page 2023

  5. [5]

    C.-H., Hickmann, K

    Biggerstaff, M., Alper, D., Dredze, M., Fox, S., Fung, I. C.-H., Hickmann, K. S., Lewis, B., Rosenfeld, R., Shaman, J., Tsou, M.-H., et al. (2016). Results from the Centers for Disease Control and Prevention’s predict the 2013--2014 Influenza Season Challenge . BMC Infectious Diseases , 16(1):1--10

    work page 2016

  6. [6]

    Billio, M., Casarin, R., Ravazzolo, F., and Van Dijk, H. K. (2013). Time-varying combinations of predictive densities using nonlinear filtering. Journal of Econometrics , 177(2):213--232

    work page 2013

  7. [7]

    G., Holmes, C

    Bissiri, P. G., Holmes, C. C., and Walker, S. G. (2016). A general framework for updating belief distributions. Journal of the Royal Statistical Society Series B: Statistical Methodology , 78(5):1103--1130

    work page 2016

  8. [8]

    Boos, D. D. and Stefanski, L. A. (2013). Essential Statistical Inference: Theory and Methods . Springer, New York

    work page 2013

Show all 80 references
  1. [9]

    L., Gneiting, T., and Reich, N

    Bracher, J., Ray, E. L., Gneiting, T., and Reich, N. G. (2021). Evaluating epidemic forecasts in an interval format. PLOS Computational Biology , 17(2):e1010592

    2021

  2. [10]

    Centers for Disease Control and Prevention FluSight: Flu Forecasting

    CDC (2024). Centers for Disease Control and Prevention FluSight: Flu Forecasting . https://www.cdc.gov/flu-forecasting/data-vis/index.html. Accessed: 2024-09-24

    2024

  3. [11]

    R., Vasnev, A

    Claeskens, G., Magnus, J. R., Vasnev, A. L., and Wang, W. (2016). The forecast combination puzzle: A simple theoretical explanation. International Journal of Forecasting , 32(3):754--762

    2016

  4. [12]

    Clemen, R. T. (1989). Combining forecasts: A review and annotated bibliography. International Journal of Forecasting , 5(4):559--583

    1989

  5. [13]

    and Iversen, E

    Clyde, M. and Iversen, E. S. (2013). Bayesian model averaging in the M-open framework . In Damien, P., Dellaportas, P., Polson, N. G., and Stephens, D. A., editors, Bayesian Theory and Applications , page 483–498. Oxford University Press

    2013

  6. [14]

    Collins, M. (2007). Ensembles and probabilities: a new era in the prediction of climate change. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences , 365(1857):1957--1970

    2007

  7. [15]

    Y., Huang, Y., Wang, Y., Ray, E

    Cramer, E. Y., Huang, Y., Wang, Y., Ray, E. L., Cornell, M., Bracher, J., Brennen, A., Castro Rivadeneira, A. J., Gerding, A., House, K., Jayawardena, D., Kanji, A. H., Khandelwal, A., Le, K., Niemi, J., Stark, A., Shah, A., Wattanachit, N., Zorn, M. W., Reich, N. G., and Cons...

    2022

  8. [16]

    C., Brooks, L

    Farrow, D. C., Brooks, L. C., Hyun, S., Tibshirani, R. J., Burke, D. S., and Rosenfeld, R. (2017). A human judgment approach to epidemiological forecasting. PLOS Computational Biology , 13(3):e1005248

    2017

  9. [17]

    J., Kim, M., Meyers, L

    Fox, S. J., Kim, M., Meyers, L. A., Reich, N. G., and Ray, E. L. (2024). Optimizing disease outbreak forecast ensembles. Emerging Infectious Diseases , 30(9):1967

    2024

  10. [18]

    T., Covey, R., Martin, G

    Frazier, D. T., Covey, R., Martin, G. M., and Poskitt, D. (2023). Solving the forecast combination puzzle. arXiv preprint arXiv:2308.05263

    2023 arXiv

  11. [19]

    Gabry, J., Češnovar, R., and Johnson, A. (2022). cmdstanr: R Interface to 'CmdStan' . https://mc-stan.org/cmdstanr/, https://discourse.mc-stan.org

    2022

  12. [20]

    W., Mayr, G

    Gebetsberger, M., Messner, J. W., Mayr, G. J., and Zeileis, A. (2018). Estimation methods for nonhomogeneous regression models: Minimum continuous ranked probability score versus maximum likelihood. Monthly Weather Review , 146(12):4323--4338

    2018

  13. [21]

    B., Stern, H

    Gelman, A., Carlin, J. B., Stern, H. S., Dunson, D. B., Vehtari, A., and Rubin, D. B. (2013). Bayesian Data Analysis, Third Edition . Chapman and Hall/CRC

    2013

  14. [22]

    and Amisano, G

    Geweke, J. and Amisano, G. (2011). Optimal prediction pools. Journal of Econometrics , 164(1):130--141

    2011

  15. [23]

    FluSight-forecast-hub

    Github (2024). FluSight-forecast-hub . https://github.com/cdcepi/FluSight-forecast-hub. Accessed: 2024-10-22

    2024

  16. [24]

    Giummol \`e , F., Mameli, V., Ruli, E., and Ventura, L. (2019). Objective bayesian inference with proper scoring rules. Test , 28(3):728--755

    2019

  17. [25]

    Gneiting, T., Balabdaoui, F., and Raftery, A. E. (2007). Probabilistic forecasts, calibration and sharpness. Journal of the Royal Statistical Society Series B: Statistical Methodology , 69(2):243--268

    2007

  18. [26]

    and Katzfuss, M

    Gneiting, T. and Katzfuss, M. (2014). Probabilistic forecasting. Annual Review of Statistics and Its Application , 1:125--151

    2014

  19. [27]

    and Raftery, A

    Gneiting, T. and Raftery, A. E. (2007). Strictly proper scoring rules, prediction, and estimation. Journal of the American Statistical Association , 102(477):359--378

    2007

  20. [28]

    E., Westveld, A

    Gneiting, T., Raftery, A. E., Westveld, A. H., and Goldman, T. (2005). Calibrated probabilistic forecasting using ensemble model output statistics and minimum CRPS estimation . Monthly Weather Review , 133(5):1098--1118

    2005

  21. [29]

    and Ranjan, R

    Gneiting, T. and Ranjan, R. (2011). Comparing density forecasts using threshold-and quantile-weighted scoring rules. Journal of Business & Economic Statistics , 29(3):411--422

    2011

  22. [30]

    and Ranjan, R

    Gneiting, T. and Ranjan, R. (2013). Combining predictive distributions. Electronic Journal of Statistics , 7:1747

    2013

  23. [31]

    A., Ngare, P., and Ikpe, D

    Gyamerah, S. A., Ngare, P., and Ikpe, D. (2020). Probabilistic forecasting of crop yields via quantile random forest and Epanechnikov kernel function . Agricultural and Forest Meteorology , 280:107808

    2020

  24. [32]

    Hall, S. G. and Mitchell, J. (2007). Combining density forecasts. International Journal of Forecasting , 23(1):1--13

    2007

  25. [33]

    Hamill, T. M. (2001). Interpretation of rank histograms for verifying ensemble forecasts. Monthly Weather Review , 129(3):550--560

    2001

  26. [34]

    Hong, T., Pinson, P., Fan, S., Zareipour, H., Troccoli, A., and Hyndman, R. J. (2016). Probabilistic energy forecasting: Global energy forecasting competition 2014 and beyond. International Journal of Forecasting , 32(3):896--913

    2016

  27. [35]

    Hong, T., Pinson, P., Wang, Y., Weron, R., Yang, D., and Zareipour, H. (2020). Energy forecasting: A review and outlook. IEEE Open Access Journal of Power and Energy , 7:376--388

    2020

  28. [36]

    Hyndman, R. J. (2020). A brief history of forecasting competitions. International Journal of Forecasting , 36(1):7--14

    2020

  29. [37]

    Hyndman, R. J. and Khandakar, Y. (2008). Automatic time series forecasting: the forecast package for R . Journal of Statistical Software , 27(3):1--22

    2008

  30. [38]

    and Tanner, M

    Jiang, W. and Tanner, M. A. (2008). Gibbs posterior for variable selection in high-dimensional classification and data mining. The Annals of Statistics , 36(5):2207--2231

    2008

  31. [39]

    Joslyn, S. L. and LeClerc, J. E. (2012). Uncertainty forecasts improve weather-related decisions and attenuate the effects of forecast error. Journal of Experimental Psychology: Applied , 18(1):126

    2012

  32. [40]

    Kapetanios, G., Mitchell, J., Price, S., and Fawcett, N. (2015). Generalised density forecast combinations. Journal of Econometrics , 188(1):150--165

    2015

  33. [41]

    Kermack, W. O. and McKendrick, A. G. (1927). A contribution to the mathematical theory of epidemics. Proceedings of the Royal Society of London. Series A, Containing papers of a mathematical and physical character , 115(772):700--721

    1927

  34. [42]

    and Korobilis, D

    Koop, G. and Korobilis, D. (2013). Large time-varying parameter VARs . Journal of Econometrics , 177(2):185--198

    2013

  35. [43]

    Lavine, I., Lindon, M., and West, M. (2021). Adaptive variable selection for sequential prediction in multivariate dynamic models. Bayesian Analysis , 16(4):1059--1083

    2021

  36. [44]

    Li, L., Kang, Y., and Li, F. (2023). Bayesian forecast combination using time-varying features. International Journal of Forecasting , 39(3):1287--1302

    2023

  37. [45]

    Li, T., Wang, Y., and Zhang, N. (2019). Combining probability density forecasts for power electrical loads. IEEE Transactions on Smart Grid , 11(2):1679--1690

    2019

  38. [46]

    C., Grushka-Cockayne, Y., and Winkler, R

    Lichtendahl Jr, K. C., Grushka-Cockayne, Y., and Winkler, R. L. (2013). Is it better to average probabilities or quantiles? Management Science , 59(7):1594--1611

    2013

  39. [47]

    M., and Frazier, D

    Loaiza-Maya, R., Martin, G. M., and Frazier, D. T. (2021). Focused Bayesian prediction . Journal of Applied Econometrics , 36(5):517--543

    2021

  40. [48]

    Makridakis, S., Spiliotis, E., and Assimakopoulos, V. (2020). The M4 competition: 100,000 time series and 61 forecasting methods. International Journal of Forecasting , 36(1):54--74

    2020

  41. [49]

    and Syring, N

    Martin, R. and Syring, N. (2022). Direct gibbs posterior inference on risk minimizers: Construction, concentration, and calibration. In Handbook of Statistics , volume 47, pages 1--41. Elsevier

    2022

  42. [50]

    M., Webber, A

    Mathis, S. M., Webber, A. E., Le \'o n, T. M., Murray, E. L., Sun, M., White, L. A., Brooks, L. C., Green, A., Hu, A. J., Rosenfeld, R., et al. (2024). Evaluation of FluSight influenza forecasting in the 2021--22 and 2022--23 seasons with a new target laboratory-confirmed infl...

    2024

  43. [51]

    and West, M

    McAlinn, K. and West, M. (2019). Dynamic Bayesian predictive synthesis in time series forecasting . Journal of Econometrics , 210(1):155--169

    2019

  44. [52]

    and Reich, N

    McAndrew, T. and Reich, N. G. (2021). Adaptively stacking ensembles for influenza forecasting. Statistics in Medicine , 40(30):6931--6952

    2021

  45. [53]

    J., Wilson, O

    Morgan, J. J., Wilson, O. C., and Menon, P. G. (2018). The wisdom of crowds approach to influenza-rate forecasting. In ASME international mechanical engineering congress and exposition , volume 52026, page V003T04A048. American Society of Mechanical Engineers

    2018

  46. [54]

    Newey, W. K. and McFadden, D. (1994). Large sample estimation and hypothesis testing. Handbook of Econometrics , 4:2111--2245

    1994

  47. [55]

    Osthus, D., Gattiker, J., Priedhorsky, R., and Del Valle, S. Y. (2019). Dynamic Bayesian influenza forecasting in the United States with hierarchical discrepancy (with discussion) . Bayesian Analysis , 14(1):261--312

    2019

  48. [56]

    and Moran, K

    Osthus, D. and Moran, K. R. (2021). Multiscale influenza forecasting. Nature Communications , 12(1):2991

    2021

  49. [57]

    Palmer, T. N. (2002). The economic value of ensemble forecasts as a tool for risk assessment: From days to decades. Quarterly Journal of the Royal Meteorological Society: A journal of the atmospheric sciences, applied meteorology and physical oceanography , 128(581):747--774

    2002

  50. [58]

    Raftery, A. E. (1996). Hypothesis testing and model selection. In Gilks, W., Richardson, S., and Spiegelhalter, D., editors, Markov Chain Monte Carlo in Practice: Interdisciplinary Statistics , pages 163--187. Chapman and Hall

    1996

  51. [59]

    E., K \'a rn \`y , M., and Ettler, P

    Raftery, A. E., K \'a rn \`y , M., and Ettler, P. (2010). Online prediction under model uncertainty via dynamic model averaging: Application to a cold rolling mill. Technometrics , 52(1):52--66

    2010

  52. [60]

    H., Van Andel, S

    Ramos, M. H., Van Andel, S. J., and Pappenberger, F. (2013). Do probabilistic forecasts lead to better decisions? Hydrology and Earth System Sciences , 17(6):2219--2232

    2013

  53. [61]

    and Gneiting, T

    Ranjan, R. and Gneiting, T. (2010). Combining probability forecasts. Journal of the Royal Statistical Society Series B: Statistical Methodology , 72(1):71--91

    2010

  54. [62]

    G., Brooks, L

    Reich, N. G., Brooks, L. C., Fox, S. J., Kandula, S., McGowan, C. J., Moore, E., Osthus, D., Ray, E. L., Tushar, A., Yamana, T. K., et al. (2019a). A collaborative multiyear, multimodel assessment of seasonal influenza forecasting in the united states. Proceedings of the Natio...

  55. [63]

    G., McGowan, C

    Reich, N. G., McGowan, C. J., Yamana, T. K., Tushar, A., Ray, E. L., Osthus, D., Kandula, S., Brooks, L. C., Crawford-Crudell, W., Gibson, G. C., et al. (2019b). A collaborative multi-model ensemble for real-time influenza season forecasting in the us. bioRxiv , page 566604

  56. [64]

    and Wallis, K

    Smith, J. and Wallis, K. F. (2009). A simple explanation of the forecast combination puzzle. Oxford Bulletin of Economics and Statistics , 71(3):331--355

    2009

  57. [65]

    Stan modeling language users guide and reference manual, 2.34

    Stan Development Team (2024). Stan modeling language users guide and reference manual, 2.34. https://mc-stan.org. Accessed: 2024-10-22

    2024

  58. [66]

    Stone, M. (1961). The opinion pool. The Annals of Mathematical Statistics , 32:1339--1342

    1961

  59. [67]

    and Martin, R

    Syring, N. and Martin, R. (2017). Gibbs posterior inference on the minimum clinically important difference. Journal of Statistical Planning and Inference , 187:67--77

    2017

  60. [68]

    and West, M

    Tallman, E. and West, M. (2024). Bayesian predictive decision synthesis. Journal of the Royal Statistical Society Series B: Statistical Methodology , 86(2):340--363

    2024

  61. [69]

    Thorey, J., Mallet, V., and Baudin, P. (2017). Online learning with the continuous ranked probability score for ensemble forecasting. Quarterly Journal of the Royal Meteorological Society , 143(702):521--529

    2017

  62. [70]

    Ulloa, N. (2019). Bayesian hierarchical modeling for disease outbreaks . PhD thesis, Iowa State University Department of Statistics

    2019

  63. [71]

    Van der Vaart, A. W. (1998). Asymptotic Statistics . Cambridge University Press, Cambridge, United Kingdom

    1998

  64. [72]

    Vehtari, A., Gelman, A., Simpson, D., Carpenter, B., and B \"u rkner, P.-C. (2021). Rank-normalization, folding, and localization: An improved \^R for assessing convergence of MCMC (with discussion) . Bayesian Analysis , 16(2):667--718

    2021

  65. [73]

    and Niemi, J

    Wadsworth, S. and Niemi, J. (2025a). Forecasting influenza hospitalizations using a bayesian hierarchical nonlinear model with discrepancy. arXiv preprint arXiv:2412.14339

    arXiv

  66. [74]

    and Niemi, J

    Wadsworth, S. and Niemi, J. (2025b). Quantile forecast matching with a Bayesian quantile Gaussian process model . arXiv preprint arXiv:2502.06605

  67. [75]

    Wadsworth, S., Niemi, J., and Reich, N. (2023). Mixture distributions for probabilistic forecasts of disease outbreaks. arXiv preprint arXiv:2310.11939

    2023 arXiv

  68. [76]

    J., Li, F., and Kang, Y

    Wang, X., Hyndman, R. J., Li, F., and Kang, Y. (2023). Forecast combinations: An over 50-year review. International Journal of Forecasting , 39(4):1518--1547

    2023

  69. [77]

    Widgren, S., Bauer, P., Eriksson, R., and Engblom, S. (2019). SimInf : An R package for data-driven stochastic disease spread simulations. Journal of Statistical Software , 91(12):1--42

    2019

  70. [78]

    Yao, Y., Vehtari, A., Simpson, D., and Gelman, A. (2018). Using stacking to average Bayesian predictive distributions (with discussion) . Bayesian Analysis , 13(3):917--1003

    2018

  71. [79]

    and Naveau, P

    Zamo, M. and Naveau, P. (2018). Estimation of the continuous ranked probability score with limited information and applications to ensemble weather forecasts. Mathematical Geosciences , 50(2):209--234

    2018

  72. [80]

    Zhang, T. (2006). Information-theoretic upper and lower bounds for statistical estimation. IEEE Transactions on Information Theory , 52(4):1307--1321

    2006

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.