Pith. sign in

REVIEW 3 major objections 4 minor 49 references

Contribution of expert aggregation to temperature prediction part i

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Online expert aggregation matches the best fixed model blend for temperature forecasts.

desk verdict The paper's oracle-matching claim is undermined by a table inconsistency, but the study is worth a serious referee. read the letter →

arxiv 2506.15217 v3 pith:ZLQYP3ES submitted 2025-06-18 math.OC

Léo Pfitzner (GMAP) , Olivier Wintenberger (SU) , Olivier Mestre (GMAP) , Marion Riverain This is my paper · ORCID
classification math.OC
keywords expertaggregationtemperatureforecastingnumericalweatherpredictiononlinelearningmodeloutputstatisticsgradienttricksecond-orderregretboundsslidingwindow
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 claims that online expert aggregation—simple algorithms that reweight weather-model forecasts as observations arrive—can produce deterministic two-metre temperature forecasts that beat every individual numerical weather prediction model, including post-processed ones, across 33 stations and lead times from 6 to 96 hours. With the gradient trick and five quantiles of the post-processed PEARP ensemble added as experts, EWA, BOA, MLprod, and MLpol reach root mean squared errors of 1.24 to 1.25 degrees Celsius, compared with 1.37 for the best single expert and 1.40 to 1.41 for a uniform average. BOA and MLpol attain the same RMSE as the best fixed convex combination of experts chosen in hindsight (1.24), without ever knowing that combination in advance. The paper also reports that the aggregations outperform the National Blend of Models and that a sliding window trades a small loss of average accuracy for greater reactivity. It acknowledges a limitation: the aggregations are too slow to switch weight to biased quantile experts during brief anomalous events such as the December 2022 cold spell at Chamonix.

What carries the argument

The machinery is convex online aggregation with regret guarantees. At each step the forecast is $\hat y_t = \sum_i w_{i,t}x_{i,t}$ with weights in the simplex, updated only from past losses, and the square loss $\ell(x,y)=(x-y)^2$ is used throughout. The gradient trick replaces the loss by its linearization $\nabla\ell_t(\hat y_t)\cdot q = \ell_t'(\hat y_t)\sum_i q_i x_{i,t}$, letting the aggregations compete against any fixed convex combination of experts rather than only the best single expert. The second-order algorithms (BOA, MLpol, MLprod) use the excess loss $\ell_{i,t}^{\mathrm{exc}} = \ell_t(\delta_i)-\sum_j w_{j,t}\ell_t(\delta_j)$ and the variance of past losses to set per-expert learning rates, which is what makes them adaptive without knowing $T$ or loss bounds in advance. A squared-error decomposition into the experts' mean error minus their diversity motivates adding the biased PEARP quantile experts.

What would settle it

Re-run the four aggregations on a temporal holdout (for instance, the final year of the dataset) or bootstrap the RMSE distribution across stations and lead times: if BOA and MLpol no longer tie the best fixed convex combination within sampling error, or if the aggregations no longer beat the best single expert, the central claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that online expert aggregation, a framework designed for worst-case adversarial prediction, transfers to deterministic temperature forecasting, even when the experts are already post-processed forecasts whose errors are nearly random. On a dataset of 33 stations in France from March 2020 to September 2023 at lead times 6 to 96 hours, the paper reports that EWA, BOA, MLprod, and MLpol, run with the gradient trick and five post-processed PEARP quantiles as additional experts, all achieve RMSEs between 1.24 and 1.25 degrees Celsius. That is below the best single expert (1.37), the uniform average (1.40–1.41), and the National Blend of Models baseline; BOA and MLpol land exactly on the RMSE of the best fixed convex combination of experts computed in hindsight, 1.24, without knowing that combination during the forecast. The paper further argues that second-order aggregations are better suited than EWA to this non-adversarial setting, and that adding deliberately biased quantile experts helps the hindsight oracles but is hard for the online rules to exploit in real time.

Load-bearing premise

The load-bearing premise is that one pooled RMSE number per method, computed on the same data used to define the oracles, is a stable basis for comparing forecast methods; without confidence intervals or a temporal holdout, the small reported gaps and the equality with the hindsight-optimal mix could be in-sample artifacts.

Editorial extensions

If this is right

  • Operational temperature guidance can be improved by roughly 0.12 to 0.13 degrees Celsius RMSE over the best single model simply by running an online reweighting scheme on forecasts that already exist.
  • Because the aggregations are online and adaptive, they can keep working across NWP model upgrades without retraining, which the paper notes is a weakness of neural-network post-processing.
  • With the gradient trick and quantile experts, BOA and MLpol recover the RMSE of the hindsight-optimal fixed convex combination (1.24), so the oracle is reachable without knowing the optimal weights in advance.
  • For short evaluation periods (under roughly 250 iterations) BOA and MLprod are the fastest learners; for long periods BOA is the most consistent of the four strategies.
  • Sliding windows make the weights visibly more reactive, but for BOA and MLpol they cost about 0.01 degrees Celsius in RMSE (1.25 vs 1.24), so reactivity and average score are in tension.

Reading between the lines

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

  • If the equality between BOA and MLpol and the hindsight-optimal combination survives a genuine out-of-sample test, it would suggest that static weight optimization has little to add over online rules for this class of forecasts.
  • Because the aggregation rules require only a convex loss, a testable extension is probabilistic aggregation with a proper scoring rule such as the continuous ranked probability score (CRPS); the paper demonstrates deterministic square loss only.
  • The Chamonix December 2022 episode shows a concrete failure mode: cumulative-loss weights are too slow to promote experts that are good only during a short cold spell. A direct test of tracking-the-best-sequence or sleeping-expert aggregations would show whether reactivity can be restored without the sliding-window RMSE penalty.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes the use of online expert aggregation algorithms (EWA, BOA, MLprod, MLpol) for deterministic 2-meter temperature forecasting, combining raw and post-processed NWP forecasts (AROME, ARPEGE, IFS/CEP) with quantiles of the post-processed PEARP ensemble. Experiments are run on 33 French stations and lead times from 6 to 96 hours, with one aggregation per station and lead time. The paper reports RMSE improvements over the best single expert and the uniform average, and claims that, when the gradient trick is used, BOA and MLpol attain the same RMSE as the best fixed convex combination of experts in hindsight. A sliding-window variant is also studied to improve reactivity to biased experts.

Significance. If the central claim holds, the paper makes a useful applied contribution: it demonstrates that online, adaptive aggregation can reproduce the hindsight-optimal fixed mix of operational deterministic temperature forecasts without knowing that mix in advance, and that this is achieved with simple algorithms and no retraining. The paper has clear strengths: it uses a real multi-station, multi-lead-time operational dataset; it compares several state-of-the-art aggregation strategies; it provides a detailed discussion of their regret bounds and practical behavior; and it makes the data available on GitHub. The main caveat is that the empirical support is currently only in the form of two-decimal point estimates, and one of the key tables appears internally inconsistent, so the headline equality between the online algorithms and the oracle is not yet established.

major comments (3)
  1. [Section 3, Tables 1–3 and Figure 8] The rows for the best convex combination and the uniform aggregation are not invariant under the gradient trick, although they must be. Table 2 (without the gradient trick, with PEARP quantiles) reports a best convex combination of 1.23°C and a uniform aggregation of 1.41°C; Table 3 (with the gradient trick, with PEARP quantiles) reports 1.24°C and 1.40°C for the same expert set. The best fixed convex combination minimizes cumulative squared error over the simplex and is independent of how the aggregation weights are updated, and the uniform aggregation is the fixed weight vector 1/N, so neither value can change between the two tables. If the oracle in Table 3 was instead computed using the linearized loss from the gradient trick, that is a different benchmark and must be stated explicitly. Otherwise one of the two tables is incorrect. Since the paper's central claim that BOA and MLpol 'have the same RMSE as this oracle' rests on the 1.24°C equality in Table 3, this inconsistency must be resolved before the claim can be evaluated.
  2. [Abstract] All RMSE comparisons are point estimates pooled over stations and lead times, reported to two decimal places, with no confidence intervals, standard errors, or significance tests. Differences of 0.01°C (for example, 1.25 vs. 1.24) are interpreted as meaningful, and the text states that 'BOA has always been the best RMSE for all sliding windows' on the basis of such differences. The authors should add uncertainty quantification, for example by bootstrap over stations and lead times or by paired forecast-error tests. They should also state explicitly that the best convex combination oracle is computed on the same evaluation period used for the online algorithms; if so, the equality with BOA and MLpol should be framed as a comparison with an in-sample hindsight benchmark rather than as an out-of-sample claim.
  3. [Section 3, Data and Figure 3] The abstract claims that the proposed methods 'show that they outperform the National Blend of Models,' but the National Blend of Models is never defined, described, or analyzed anywhere in Sections 3 or 4, and no corresponding row or table is provided. This claim is therefore unsupported by the manuscript as written. The authors should either add the NBM comparison and describe the data used for it, or remove the claim from the abstract and introduction.
minor comments (4)
  1. [Section 3, Data and Figure 3] The lead-time description is inconsistent: the text lists the additional lead times as 57, 72 and 96 hours, while Figures 3, 5 and 6 and the discussion use an 84-hour lead time. Please align the data description with the figures and the text.
  2. [Figure captions] The captions of Figure 7 and Figure 9 contain the typo 'Weigths' instead of 'Weights', and the caption of Table 1 says 'MLpod' instead of 'MLprod'.
  3. [Section 3, Sliding Window] The notation for the sliding window is ambiguous: replacing a cumulative sum $X_t = \sum_{s=1}^t x_s$ by $X_t = \sum_{s=t-w}^t x_s$ is understandable, but the accompanying text 'each variable of the form $X_t = \min(x_s, 1 \le s \le t)$ by $X_t = \min(x_s, t-w \le s \le t)$' is not written in standard notation and should be clarified.
  4. [General presentation] The manuscript contains duplicated passages: the discussion of the Chamonix weights appears twice around Figure 7 and Figure 9, and the conclusion text is repeated near the end of the file. These duplicates should be removed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the evaluation compares online algorithms to hindsight oracles, and no fitted parameter is renamed as a prediction.

full rationale

The paper's derivation chain is self-contained and not circular. The aggregation algorithms EWA, BOA, MLprod, and MLpol are run online with adaptive learning rates, and the RMSE values in Tables 1-3 are cumulative losses over the test sequence rather than fitted constants. The 'best expert' and 'best convex combination' oracles are hindsight benchmarks computed from the same expert predictions and observations, but they are not inputs to the algorithms; comparing online performance to a hindsight oracle is the standard, non-circular evaluation protocol for expert aggregation. The only self-citations are to Wintenberger's published BOA regret-bound work and to co-authored post-processing references, which are external published results rather than assumptions that already contain the temperature-prediction outcome. The inconsistency between Tables 2 and 3 in the oracle and uniform-aggregation RMSEs (1.23 vs 1.24 and 1.41 vs 1.40) is a numerical or reporting concern that the authors should resolve, but it does not exhibit a reduction of the claimed prediction to its inputs: even if the oracle were redefined to a linearized-loss oracle, the paper would need to state that, and the algorithms' updates do not use the oracle. No circular step is exhibited.

Assumptions & free parameters 1 free parameters · 3 assumptions · 1 invented entities

The central claim rests on published regret-bound theorems and on a domain assumption about the stochastic nature of the atmosphere. The only hand-chosen hyperparameter, the sliding window size, is scanned rather than fitted, and it does not drive the main RMSE results in Tables 1 to 3.

free parameters (1)
  • sliding window size w = 60 to 1253 days, scanned
    Section 3 varies the sliding window size by hand to study reactivity. The main RMSE tables do not depend on a single fitted value, so this hyperparameter is not load-bearing for the central claim.
assumptions (3)
  • standard math The regret bounds for EWA, MLprod, MLpol and BOA quoted in Section 2 are correct and applicable.
    The paper imports these bounds from Cesa-Bianchi and Lugosi, Gaillard et al., and Wintenberger; they are not re-derived in this preprint.
  • domain assumption Losses used by MLprod and MLpol satisfy the boundedness conditions required by their regret bounds.
    Equations 11 to 14 assume losses in [0,1], while squared temperature errors are not formally bounded in the experiment. The paper uses adaptive learning rates without verifying this condition.
  • domain assumption The atmosphere behaves more like a stochastic process than an adversary.
    Section 2 uses this claim to motivate the second-order algorithms as better suited to weather forecasting. It is a plausible modeling choice but is not directly tested.
invented entities (1)
  • None
    purpose: No new particles, mediators, forces, dimensions, or conserved quantities are introduced.
    The PEARP quantiles are existing forecast products used as experts, not invented entities. The paper contains no new ontological commitments.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Contribution of expert aggregation to temperature prediction part i." pith.science (2026). https://pith.science/paper/ZLQYP3ES

@misc{pith2026250615217,
  author       = {Pith},
  title        = {Pith review of: Contribution of expert aggregation to temperature prediction part i},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZLQYP3ES}},
  note         = {Machine review of arXiv:2506.15217}
}
read the original abstract

Many Numerical Weather Prediction models and their associated Post-Processed Models are available. Combining all of these predictions in an optimal way is however not straightforward. This can be achieved thanks to Expert Aggregation (EA) which has many advantages, such as being online, being adaptive to model changes and having theoretical guarantees. In this paper, we propose a method for making deterministic temperature predictions with EA. We used Exponentially Weighted Average, MLprod and MLpol and Bernstein Online Aggregation. Hence, we combine and outperform the forecasts of the raw and post-processed Integrated Forecasting System (IFS), forecasts of Application of Research to Operations at Mesoscale (AROME), Action de Recherche Petite Echelle Grande Echelle (ARPEGE) and quantiles of the post processed Pr{\'e}vision d'Ensemble ARPEGE (PEARP). We also compare the different EA strategies in various settings and show that they outperform the National Blend of Models. Finally, we discuss certain limitations.

Figures

Figures reproduced from arXiv: 2506.15217 by the authors.

Figure 1
Figure 1. Monthly RMSE for 2 meters above ground temperature predictions (midnight run lead time 36h) of different [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Location of the weather stations. Can we beat the best expert in hindsight ? First of all, with the help of the Opera library [Gaillard and Goude, 2016] which we used in the entire study, we simply run one EWA, one BOA one MLprod and one MLpol aggregation without the gradient trick and without the PEARP quantiles for each lead time and station. We then compared these aggregations by computing the Root Mean Squared E… view at source ↗
Figure 3
Figure 3. Box plot of the station’s RMSE for different lead times and aggregations (without the gradient trick and [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Observed temperatures in black and predicted temperatures for the Chamonix station (French Alps) for the [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Box plot of the station’s RMSE for different lead times and aggregations (without the gradient trick and with [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Box plot of the station’s RMSE for different lead times and aggregations (with the gradient trick and the [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 8
Figure 8. Figure 8: Weigths of the experts, for EWA, BOA, MLpol and MLprod with the g for 1253 iterations from 2020-03-30 to 2023-09-03, in Chamonix for the lead time 48 hours. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 8
Figure 8. Figure 8: RMSE over all the stations and lead times of the expert aggregations (with the gradient trick and the PEARP [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 10
Figure 10. Figure 10: Weigths of the experts, for BOA MLpol MLprod and EWA (with the gradient trick and and a sliding window of 500 days), for 1253 iterations from 2020-03-30 to 2023-09-03, in Chamonix for the lead time [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 20 canonical work pages

  1. [1]

    Prediction, Learning , and Games

    Nicolo Cesa-Bianchi and Gabor Lugosi. Prediction, Learning , and Games . Cambridge University Press, March 2006. ISBN 978-1-139-45482-7. Google-Books-ID: zDnRBlazhfYC

    2006

  2. [2]

    A Second -order Bound with Excess Losses

    Pierre Gaillard, Gilles Stoltz, and Tim van Erven. A Second -order Bound with Excess Losses . Conference on Learning Theory. PMLR., page 21, 2014

    2014

  3. [3]

    Stochastic online convex optimization

    Olivier Wintenberger. Stochastic online convex optimization. Application to probabilistic time series forecasting. Electronic Journal of Statistics, 18 0 (1): 0 429--464, January 2024. ISSN 1935-7524, 1935-7524. doi:10.1214/23-EJS2208. URL https://projecteuclid.org/journals/electronic-journal-of-statistics/volume-18/issue-1/Stochastic-online-convex-optimi...

    doi:10.1214/23-ejs2208 2024

  4. [4]

    Seity, P

    Y. Seity, P. Brousseau, S. Malardel, G. Hello, P. Bénard, F. Bouttier, C. Lac, and V. Masson. The AROME - France Convective - Scale Operational Model . Monthly Weather Review, 139 0 (3): 0 976--991, March 2011. ISSN 1520-0493, 0027-0644. doi:10.1175/2010MWR3425.1. URL https://journals.ametsoc.org/view/journals/mwre/139/3/2010mwr3425.1.xml. Publisher: Amer...

    doi:10.1175/2010mwr3425.1 2011

  5. [5]

    Operational Convective - Scale Numerical Weather Prediction with the COSMO Model : Description and Sensitivities

    Michael Baldauf, Axel Seifert, Jochen Förstner, Detlev Majewski, Matthias Raschendorfer, and Thorsten Reinhardt. Operational Convective - Scale Numerical Weather Prediction with the COSMO Model : Description and Sensitivities . Monthly Weather Review, 139 0 (12): 0 3887--3905, December 2011. ISSN 1520-0493, 0027-0644. doi:10.1175/MWR-D-10-05013.1. URL htt...

    doi:10.1175/mwr-d-10-05013.1 2011

  6. [6]

    Déqué, C

    M. Déqué, C. Dreveton, A. Braun, and D. Cariolle. The ARPEGE / IFS atmosphere model: a contribution to the French community climate modelling. Climate Dynamics, 10 0 (4): 0 249--266, September 1994. ISSN 1432-0894. doi:10.1007/BF00208992. URL https://doi.org/10.1007/BF00208992

    work page doi:10.1007/bf00208992 1994

  7. [7]

    The quiet revolution of numerical weather prediction

    Peter Bauer, Alan Thorpe, and Gilbert Brunet. The quiet revolution of numerical weather prediction. Nature, 525 0 (7567): 0 47--55, September 2015. ISSN 1476-4687. doi:10.1038/nature14956. URL https://www.nature.com/articles/nature14956. Number: 7567 Publisher: Nature Publishing Group

    doi:10.1038/nature14956 2015

  8. [8]

    Descamps, C

    L. Descamps, C. Labadie, A. Joly, E. Bazile, P. Arbogast, and P. Cébron. PEARP , the Météo - France short-range ensemble prediction system. Quarterly Journal of the Royal Meteorological Society, 141 0 (690): 0 1671--1685, 2015. ISSN 1477-870X. doi:10.1002/qj.2469. URL https://onlinelibrary.wiley.com/doi/abs/10.1002/qj.2469. \_eprint: https://onlinelibrary...

    work page doi:10.1002/qj.2469 2015

Show all 49 references
  1. [9]

    Hamill and Stephen J

    Thomas M. Hamill and Stephen J. Colucci. Verification of Eta – RSM Short - Range Ensemble Forecasts . Monthly Weather Review, 125 0 (6): 0 1312--1327, June 1997. ISSN 1520-0493, 0027-0644. doi:10.1175/1520-0493(1997)125<1312:VOERSR>2.0.CO;2. URL https://journals.ametsoc.org/vi...

    1997 doi

  2. [10]

    Edward N. Lorenz. Deterministic Nonperiodic Flow . Journal of the Atmospheric Sciences, 20 0 (2): 0 130--141, March 1963. ISSN 0022-4928, 1520-0469. doi:10.1175/1520-0469(1963)020<0130:DNF>2.0.CO;2. URL https://journals.ametsoc.org/view/journals/atsc/20/2/1520-0469_1963_020_01...

    1963 doi

  3. [11]

    Holton and D

    James R. Holton and D. O. Staley. An Introduction to Dynamic Meteorology . American Journal of Physics, 41 0 (5): 0 752--754, 2013. ISSN 0002-9505. doi:10.1119/1.1987371. URL https://doi.org/10.1119/1.1987371

    2013 doi

  4. [12]

    Predictability of large-scale atmospheric motions: Lyapunov exponents and error dynamics

    Stéphane Vannitsem. Predictability of large-scale atmospheric motions: Lyapunov exponents and error dynamics. Chaos: An Interdisciplinary Journal of Nonlinear Science, 27 0 (3): 0 032101, March 2017. ISSN 1054-1500. doi:10.1063/1.4979042. URL https://doi.org/10.1063/1.4979042

    2017 doi

  5. [13]

    Stéphane Vannitsem, John Bjørnar Bremnes, Jonathan Demaeyer, Gavin R. Evans, Jonathan Flowerdew, Stephan Hemri, Sebastian Lerch, Nigel Roberts, Susanne Theis, Aitor Atencia, Zied Ben Bouallègue, Jonas Bhend, Markus Dabernig, Lesley De Cruz, Leila Hieta, Olivier Mestre, Lionel ...

    2021

  6. [14]

    From research to applications – examples of operational ensemble post-processing in France using machine learning

    Maxime Taillardat and Olivier Mestre. From research to applications – examples of operational ensemble post-processing in France using machine learning. Nonlinear Processes in Geophysics, 27 0 (2): 0 329--347, May 2020. ISSN 1023-5809. doi:10.5194/npg-27-329-2020. URL https://...

    2020 doi

  7. [15]

    Neural Networks for Postprocessing Ensemble Weather Forecasts

    Stephan Rasp and Sebastian Lerch. Neural Networks for Postprocessing Ensemble Weather Forecasts . Monthly Weather Review, 146 0 (11): 0 3885--3900, November 2018. ISSN 1520-0493, 0027-0644. doi:10.1175/MWR-D-18-0187.1. URL https://journals.ametsoc.org/view/journals/mwre/146/11...

    2018 doi

  8. [16]

    Pangu- Weather : A 3D High - Resolution Model for Fast and Accurate Global Weather Forecast , November 2022

    Kaifeng Bi, Lingxi Xie, Hengheng Zhang, Xin Chen, Xiaotao Gu, and Qi Tian. Pangu- Weather : A 3D High - Resolution Model for Fast and Accurate Global Weather Forecast , November 2022. URL http://arxiv.org/abs/2211.02556. arXiv:2211.02556 [physics]

    2022 arXiv

  9. [17]

    FourCastNet : A Global Data -driven High -resolution Weather Model using Adaptive Fourier Neural Operators , February 2022

    Jaideep Pathak, Shashank Subramanian, Peter Harrington, Sanjeev Raja, Ashesh Chattopadhyay, Morteza Mardani, Thorsten Kurth, David Hall, Zongyi Li, Kamyar Azizzadenesheli, Pedram Hassanzadeh, Karthik Kashinath, and Animashree Anandkumar. FourCastNet : A Global Data -driven Hig...

    2022 arXiv

  10. [18]

    GraphCast : Learning skillful medium-range global weather forecasting, December 2022

    Remi Lam, Alvaro Sanchez-Gonzalez, Matthew Willson, Peter Wirnsberger, Meire Fortunato, Alexander Pritzel, Suman Ravuri, Timo Ewalds, Ferran Alet, Zach Eaton-Rosen, Weihua Hu, Alexander Merose, Stephan Hoyer, George Holland, Jacklynn Stott, Oriol Vinyals, Shakir Mohamed, and P...

    2022 arXiv

  11. [19]

    Gupta, and Aditya Grover

    Tung Nguyen, Johannes Brandstetter, Ashish Kapoor, Jayesh K. Gupta, and Aditya Grover. ClimaX : A foundation model for weather and climate, February 2023. URL http://arxiv.org/abs/2301.10343. arXiv:2301.10343 [cs]

    2023 arXiv

  12. [20]

    FengWu : Pushing the Skillful Global Medium -range Weather Forecast beyond 10 Days Lead , April 2023

    Kang Chen, Tao Han, Junchao Gong, Lei Bai, Fenghua Ling, Jing-Jia Luo, Xi Chen, Leiming Ma, Tianning Zhang, Rui Su, Yuanzheng Ci, Bin Li, Xiaokang Yang, and Wanli Ouyang. FengWu : Pushing the Skillful Global Medium -range Weather Forecast beyond 10 Days Lead , April 2023. URL ...

    2023

  13. [21]

    Simon Lang, Mihai Alexe, Matthew Chantry, Jesper Dramsch, Florian Pinault, Baudouin Raoult, Mariana C. A. Clare, Christian Lessig, Michael Maier-Gerber, Linus Magnusson, Zied Ben Bouallègue, Ana Prieto Nemesio, Peter D. Dueben, Andrew Brown, Florian Pappenberger, and Florence ...

    2024 arXiv

  14. [22]

    Elizabeth E. Ebert. Ability of a Poor Man 's Ensemble to Predict the Probability and Distribution of Precipitation . Monthly Weather Review, 129 0 (10): 0 2461--2480, October 2001. ISSN 1520-0493, 0027-0644. doi:10.1175/1520-0493(2001)129<2461:AOAPMS>2.0.CO;2. URL https://jour...

    2001 doi

  15. [23]

    Mylne, Ruth E

    Kenneth R. Mylne, Ruth E. Evans, and Robin T. Clark. Multi-model multi-analysis ensembles in quasi-operational medium-range forecasting. Quarterly Journal of the Royal Meteorological Society, 128 0 (579): 0 361--384, 2002. ISSN 1477-870X. doi:10.1256/00359000260498923. URL htt...

    2002 doi

  16. [24]

    Agrégation séquentielle de prédicteurs: méthodologie générale et applications à la prévision de la qualité de l’air et à celle de la consommation électrique

    Gilles Stoltz. Agrégation séquentielle de prédicteurs: méthodologie générale et applications à la prévision de la qualité de l’air et à celle de la consommation électrique. Journal de la Société française de Statistique, 151 0 (2): 0 66--106, 2010

    2010

  17. [25]

    Aggregating Strategies

    Vladimir Vovk. Aggregating Strategies . Proceedings of the 3rd Annual Workshop on Computational Learning Theory, 1990

    1990

  18. [26]

    Littlestone and M

    N. Littlestone and M. K. Warmuth. The Weighted Majority Algorithm . Information and Computation, 108 0 (2): 0 212--261, February 1994. ISSN 0890-5401. doi:10.1006/inco.1994.1009. URL https://www.sciencedirect.com/science/article/pii/S0890540184710091

    1994

  19. [27]

    Koolen, and Dirk van der Hoeven

    Tim van Erven, Wouter M. Koolen, and Dirk van der Hoeven. MetaGrad : Adaptation using Multiple Learning Rates in Online Learning , August 2021. URL http://arxiv.org/abs/2102.06622. Number: arXiv:2102.06622 arXiv:2102.06622 [cs, stat]

    2021 arXiv

  20. [28]

    Second-order quantile methods for experts and combinatorial games

    Wouter M Koolen and Tim Van Erven. Second-order quantile methods for experts and combinatorial games. In Conference on Learning Theory, pages 1155--1175. PMLR, 2015

    2015

  21. [29]

    Optimal Learning with Bernstein Online Aggregation

    Olivier Wintenberger. Optimal Learning with Bernstein Online Aggregation . Machine Learning, 106 0 (1): 0 119--141, 2017

    2017

  22. [30]

    Ozone ensemble forecast with machine learning algorithms

    Vivien Mallet, Gilles Stoltz, and Boris Mauricette. Ozone ensemble forecast with machine learning algorithms. Journal of Geophysical Research: Atmospheres, 114 0 (D5), 2009. ISSN 2156-2202. doi:https://doi.org/10.1029/2008JD009978. URL https://agupubs.onlinelibrary.wiley.com/d...

    2009 doi

  23. [31]

    Forecasting electricity consumption by aggregating specialized experts

    Marie Devaine, Pierre Gaillard, Yannig Goude, and Gilles Stoltz. Forecasting electricity consumption by aggregating specialized experts. Machine Learning, 90 0 (2): 0 231--260, February 2013. ISSN 1573-0565. doi:10.1007/s10994-012-5314-7. URL https://doi.org/10.1007/s10994-012-5314-7

    2013 doi

  24. [32]

    Contributions à l’agrégation séquentielle robuste d’experts : Travaux sur l’erreur d’approximation et la prévision en loi

    Pierre Gaillard. Contributions à l’agrégation séquentielle robuste d’experts : Travaux sur l’erreur d’approximation et la prévision en loi. Applications à la prévision pour les marchés de l’énergie. phdthesis, Université Paris Sud - Paris XI, July 2015. URL https://tel.archive...

    2015

  25. [33]

    Fundamentals and exchange rate forecastability with simple machine learning methods

    Christophe Amat, Tomasz Michalski, and Gilles Stoltz. Fundamentals and exchange rate forecastability with simple machine learning methods. Journal of International Money and Finance, 88: 0 1--24, November 2018. ISSN 0261-5606. doi:10.1016/j.jimonfin.2018.06.003. URL http://www...

    2018 doi

  26. [34]

    Expert Aggregation for Financial Forecasting

    Carl Remlinger, Clémence Alasseur, Marie Brière, and Joseph Mikael. Expert Aggregation for Financial Forecasting . The Journal of Finance and Data Science, page 100108, November 2023. ISSN 2405-9188. doi:10.1016/j.jfds.2023.100108. URL https://www.sciencedirect.com/science/art...

    2023

  27. [35]

    Sequential model aggregation for production forecasting

    Raphaël Deswarte, Véronique Gervais, Gilles Stoltz, and Sébastien Da Veiga. Sequential model aggregation for production forecasting. Computational Geosciences, 23 0 (5): 0 1107--1124, October 2019. ISSN 1573-1499. doi:10.1007/s10596-019-09872-1. URL https://doi.org/10.1007/s10...

    2019 doi

  28. [36]

    Thorey, C

    J. Thorey, C. Chaussin, and V. Mallet. Ensemble forecast of photovoltaic power with online CRPS learning. International Journal of Forecasting, 34 0 (4): 0 762--773, October 2018. ISSN 0169-2070. doi:10.1016/j.ijforecast.2018.05.007. URL https://www.sciencedirect.com/science/a...

    2018 doi

  29. [37]

    CRPS -based online learning for nonlinear probabilistic forecast combination

    Dennis van der Meer, Pierre Pinson, Simon Camal, and Georges Kariniotakis. CRPS -based online learning for nonlinear probabilistic forecast combination. International Journal of Forecasting, January 2024. ISSN 0169-2070. doi:10.1016/j.ijforecast.2023.12.005. URL https://www.sc...

    2024 doi

  30. [38]

    Sequential aggregation of probabilistic forecasts— Application to wind speed ensemble forecasts

    Michaël Zamo, Bell Lilianne, and Mestre Olivier. Sequential aggregation of probabilistic forecasts— Application to wind speed ensemble forecasts. Journal of the Royal Statistical Society: Series C (Applied Statistics) - Wiley Online Library, 2021. URL https://rss.onlinelibrary...

    2021 doi

  31. [39]

    Thorey, V

    J. Thorey, V. Mallet, and P. Baudin. Online learning with the Continuous Ranked Probability Score for ensemble forecasting. Quarterly Journal of the Royal Meteorological Society, 143 0 (702): 0 521--529, 2017. ISSN 1477-870X. doi:10.1002/qj.2940. URL https://onlinelibrary.wile...

    2017 doi

  32. [40]

    Boosting diversity in regression ensembles

    Mathias Bourel, Jairo Cugliari, Yannig Goude, and Jean-Michel Poggi. Boosting diversity in regression ensembles. Statistical Analysis and Data Mining: The ASA Data Science Journal, 17 0 (1): 0 e11654, 2024. ISSN 1932-1872. doi:10.1002/sam.11654. URL https://onlinelibrary.wiley...

    2024 doi

  33. [41]

    Learning the Learning Rate for Prediction with Expert Advice

    Wouter M Koolen, Tim van Erven, and Peter Grünwald. Learning the Learning Rate for Prediction with Expert Advice . In Advances in Neural Information Processing Systems , volume 27. Curran Associates, Inc., 2014. URL https://proceedings.neurips.cc/paper/2014/hash/3f67fd97162d20...

    2014

  34. [42]

    Adaptive and Self - Confident On - Line Learning Algorithms

    Peter Auer, Nicolò Cesa-Bianchi, and Claudio Gentile. Adaptive and Self - Confident On - Line Learning Algorithms . Journal of Computer and System Sciences, 64 0 (1): 0 48--75, February 2002. ISSN 0022-0000. doi:10.1006/jcss.2001.1795. URL https://www.sciencedirect.com/science...

    2002

  35. [43]

    Improved second-order bounds for prediction with expert advice

    Nicolò Cesa-Bianchi, Yishay Mansour, and Gilles Stoltz. Improved second-order bounds for prediction with expert advice. Machine Learning, 66 0 (2): 0 321--352, March 2007. ISSN 1573-0565. doi:10.1007/s10994-006-5001-7. URL https://doi.org/10.1007/s10994-006-5001-7

    2007 doi

  36. [44]

    Kalman recursions Aggregated Online

    Eric Adjakossa, Yannig Goude, and Olivier Wintenberger. Kalman recursions Aggregated Online . Statistical Papers, March 2023. ISSN 1613-9798. doi:10.1007/s00362-023-01410-7. URL https://doi.org/10.1007/s00362-023-01410-7

    2023 doi

  37. [45]

    Progressive mixture rules are deviation suboptimal

    Jean-yves Audibert. Progressive mixture rules are deviation suboptimal. In Advances in Neural Information Processing Systems , volume 20. Curran Associates, Inc., 2007. URL https://proceedings.neurips.cc/paper/2007/hash/ef575e8837d065a1683c022d2077d342-Abstract.html

    2007

  38. [46]

    Opera: Online Prediction by Expert Aggregation , 2016

    Pierre Gaillard and Yannig Goude. Opera: Online Prediction by Expert Aggregation , 2016. URL http://pierre.gaillard.me/opera.html

    2016

  39. [47]

    Mark Herbster and Manfred K. Warmuth. Tracking the Best Expert . Machine Learning, 32 0 (2): 0 151--178, August 1998. ISSN 1573-0565. doi:10.1023/A:1007424614876. URL https://doi.org/10.1023/A:1007424614876

    1998 doi

  40. [48]

    Efficient tracking of a growing number of experts

    Jaouad Mourtada and Odalric-Ambrym Maillard. Efficient tracking of a growing number of experts. arXiv:1708.09811 [cs, stat], August 2017. URL http://arxiv.org/abs/1708.09811. arXiv: 1708.09811

    2017 arXiv

  41. [49]

    Schapire, Yoram Singer, and Manfred K

    Yoav Freund, Robert E. Schapire, Yoram Singer, and Manfred K. Warmuth. Using and combining predictors that specialize. In Proceedings of the twenty-ninth annual ACM symposium on Theory of computing , STOC '97, pages 334--343, New York, NY, USA, May 1997. Association for Comput...

    1997

Pith tools

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