Pith. sign in

REVIEW 4 major objections 4 minor 80 references

E-STGCN: Extreme Spatiotemporal Graph Convolutional Networks for Air Quality Forecasting

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read E-STGCN, a graph convolutional forecaster with an extreme-value theory loss, claims superior 30-, 60-, and 90-day air quality forecasts for Delhi's PM2.5, PM10, and NO2, with the largest gains in winter.

desk verdict A serious STGCN+EVT hybrid whose central POT loss is mis-specified, so the extreme-value claim falls as written; worth a major-revision path, not a reject. read the letter →

arxiv 2411.12258 v2 pith:XVXWWIQ4 submitted 2024-11-19 stat.AP stat.ME

Madhurima Panja , Tanujit Chakraborty , Anubhab Biswas , Soudeep Deb This is my paper · ORCID
classification stat.APstat.ME MSC 62G3262M2062P1268T07
keywords airqualityforecastingextremevaluetheorygraphconvolutionalnetworksspatiotemporalgeneralizedParetodistributionpeaksoverthresholdLSTMDelhipollution
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 proposes E-STGCN, a spatiotemporal deep learning forecaster that grafts extreme value theory onto the standard STGCN architecture by replacing temporal convolutions with LSTM units and adding a loss term built from a generalized Pareto fit to threshold exceedances. The authors aim to show that this hybrid model forecasts Delhi's daily PM2.5, PM10, and NO2 concentrations more accurately than thirteen temporal and spatiotemporal baselines, with the largest margins in winter, when pollutant levels soar. If true, the model offers a practical early-warning tool: it predicts peaks that violate India's NAAQS regulatory limits, not just average concentrations, and wraps the point forecasts in conformal prediction intervals. The paper's central bet is that embedding tail statistics into the training objective is what buys the improvement, not just the architectural upgrade to LSTM.

What carries the argument

Three modules carry the argument. The spatial module runs graph convolutions over a static adjacency matrix built from the Haversine distance between monitoring stations, using a first-order Chebyshev approximation of the graph Laplacian filter to propagate information across neighboring sites. The temporal module replaces STGCN's gated one-dimensional convolutions with an LSTM sequence model, which is meant to capture the long-memory dynamics of pollution series. The EVT module fits a generalized Pareto distribution to each station's concentrations above the NAAQS threshold (60 µg/m³ for PM2.5, 100 for PM10, 80 for NO2), estimates the shape and scale parameters by maximum likelihood via BFGS, and feeds the negative log-likelihood of the fitted GP into a hybrid loss: predictions below the threshold are trained with plain MSE, while predictions above it add β1·MSE + β2·POTL, so the optimizer balances squared-error accuracy against conformance to the learned tail shape.

What would settle it

Retrain E-STGCN and the EVT-free modified STGCN on a pollutant or city where exceedances are rare (below roughly 5% of days) and measure whether tail forecasts differ; the paper's own near-tie between the two models for NO2, at about 10% exceedance, already indicates where the method's advantage ends. In parallel, inspect the GP fit for high-exceedance stations: if the fitted shape parameter and probability plots show poor agreement, the POT loss is an arbitrary penalty rather than a statistically grounded extreme-value prior.

Watch

Extended reading notes

Core claim

The central claim, stated in Section 1, is that E-STGCN, which enhances the original STGCN by incorporating both LSTM-based temporal modeling and an EVT-based loss function, achieves superior performance while capturing extreme air pollutant concentrations. Empirically, the paper reports that E-STGCN ranks first among all tested forecasters on the MCB test for PM2.5 (mean rank 3.27 on RMSE) and PM10 (2.34), that it improves on the best baseline by 22.4% in MAE in one winter month, and that it stays competitive in summer, where exceedances are rare. The authors also note that no single model wins every task, in line with the No Free Lunch theorem, with ARIMA and NBeats taking some months. For NO2, where about 10% of observations exceed the threshold, E-STGCN and the modified STGCN without the EVT component perform nearly identically, which the authors read as confirming that the POT loss matters mainly where extremes are frequent.

Load-bearing premise

The load-bearing premise is that concentrations above the NAAQS threshold follow a generalized Pareto distribution, so the POT penalty is a statistically grounded prior; but 61% of PM2.5 days and 77% of PM10 days exceed those thresholds, making the tail the bulk of the data and straining the asymptotic basis of the GP approximation.

Editorial extensions

If this is right

  • E-STGCN provides a concrete warning-system architecture: daily forecasts up to 90 days ahead, with the largest documented gains in the winter months when PM2.5 and PM10 exceed regulatory limits most often.
  • The hybrid POT loss is model-agnostic, so the same 'EVT penalty on exceedances' recipe can be attached to any spatiotemporal forecaster, not only STGCN.
  • Combined with conformal prediction, the model outputs probabilistic intervals, giving policy users uncertainty bands around the point forecasts.
  • The study's internal ablation, modified STGCN versus E-STGCN, isolates the contribution of the EVT loss: where exceedances are common the EVT loss buys accuracy, and where they are rare it neither helps nor hurts.
  • The runtime comparison shows the EVT machinery adds little computational cost over standard STGCN, keeping the model feasible for near-real-time forecasting.

Reading between the lines

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

  • If the EVT loss is what drives the winter gains, then a simpler exceedance-weighted MSE that up-weights squared error on days above the threshold might reproduce most of the benefit without invoking asymptotic GP theory; comparing the two penalties on the same architecture would settle whether the GP structure or the up-weighting is the active ingredient.
  • The paper's own threshold analysis marks the boundary of the method: at NAAQS thresholds the exceedance fraction is the majority of the data for PM2.5 and PM10, so the GP approximation is being used far outside its asymptotic regime; the mean-excess-plot thresholds the authors compute (583 for PM2.5, 658 for PM10) define where EVT is formally defensible, and the method's performance there is untes
  • Because training uses 2019–2022 and testing uses 2023, a natural stress test is to retrain on shifted seasonal conditions or on another megacity with comparable monitoring density to see whether the winter superiority persists.
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

4 major / 4 minor

Summary. The paper proposes E-STGCN, a spatiotemporal graph convolutional network for daily air-quality forecasting at 37 Delhi monitoring stations. The architecture combines a GCN spatial module, an LSTM temporal module, and an extreme-value-theory (EVT) module that fits a generalized Pareto (GP) distribution to threshold exceedances and adds a negative log-likelihood penalty (POT loss) to the MSE objective. The authors evaluate 30-, 60-, and 90-day-ahead forecasts for PM2.5, PM10, and NO2 against 13 baselines, report MCB and Diebold-Mariano tests, and add conformal prediction intervals. The central claim is that E-STGCN achieves superior forecasting performance and better captures extreme pollutant concentrations.

Significance. If the EVT-guided mechanism were correctly specified and the empirical claims accurately stated, this would be a useful contribution: it combines a domain-relevant statistical prior with deep spatiotemporal forecasting, uses public real-world data over multiple horizons, includes a broad benchmark comparison, and ships code and data. The MCB/DM tests and conformal intervals are commendable additions. However, the paper's load-bearing EVT loss is mis-specified in the main text, the seasonal consistency claim is contradicted by the paper's own tables, and the threshold choice makes the GP tail assumption questionable for PM2.5 and PM10. These issues currently prevent the results from supporting the advertised conclusions.

major comments (4)
  1. [Section 3.2.3, Eq. (6)] The POT loss is not the negative log-likelihood of the fitted GP distribution for exceedances. Since the GP in Eq. (1) models the excess variable z = X - tau, the NLL for a predicted concentration Z_hat_i^t > tau should be evaluated at Z_hat_i^t - tau. Equation (6) instead substitutes the raw predicted value Z_hat_i^t into log sigma_hat + (1 + 1/xi_hat) log(1 + xi_hat Z_hat_i^t / sigma_hat), and Eq. (7) activates this term whenever Z_hat_i^t > tau. As written, the regularization is not the likelihood of the excess distribution, so the central claim that E-STGCN is 'EVT-guided' is not supported by the stated loss. Please correct the argument of the GP NLL to Z_hat_i^t - tau and re-evaluate, or justify why the raw-scale penalty is intended.
  2. [Section 4.2, threshold choice] The average exceedance rates reported in Section 4.2 are 61% for PM2.5 and 77% for PM10 above the NAAQS thresholds. The GP approximation in Eq. (1) is an asymptotic result for exceedances over a large threshold; with the majority of observations above the threshold, the exceedances are the bulk of the data and the GP assumption is not automatically valid. The paper should provide threshold-stability diagnostics (e.g., GP goodness-of-fit, shape-parameter stability across thresholds) or sensitivity analyses with alternative thresholds; otherwise the POT term in Eq. (7) is an ad hoc penalty rather than a statistically grounded extreme-value prior.
  3. [Abstract and Section 4.4] The abstract's claim of 'consistent performance across all seasons' is contradicted by the reported tables. For example, in Table 2, December 2023 is won by NBeats/ARIMA (MAE 68.77/74.03 vs. 116.77 for E-STGCN), October by GpGp/Transformers, and several other months are not won by E-STGCN. In Table 3, STNN wins June and NBeats wins July and December for PM10. In Table 4, STARMA wins March and June for NO2. The authors should either weaken the claim to 'competitive or best in many periods' or provide a statistical summary that justifies 'consistent performance' across seasons.
  4. [Section 3.2.1, Eq. (2)] The Haversine distance formula is written with cos(lambda_i) cos(lambda_j) in the second term, but the text defines phi as latitude and lambda as longitude. The standard Haversine formula uses cos(phi_i) cos(phi_j) sin^2(Delta_lambda / 2). If the formula as printed is what is implemented, the adjacency matrix incorrectly weights stations by longitude rather than latitude; if it is a typo, the equation should be corrected. Since the adjacency matrix is the basis of the spatial module, this needs to be fixed and confirmed.
minor comments (4)
  1. [Table 1] The column header 'Extreme V alue Handling' contains a typo; it should be 'Extreme Value Handling'.
  2. [Section 3.2.4] The text 'violate regulatory thresholds by NAAQs' should read 'NAAQS' (National Ambient Air Quality Standards).
  3. [Section 4.3] The CRPS definition states that F_i^t is 'the predicted probability distribution function,' but the integral uses F as a cumulative distribution function. Please clarify the notation to distinguish the predictive CDF from a density.
  4. [Section 4.5] When presenting the DM test for NO2, the authors note that the loss differential is near zero and the statistic is undefined; this should be stated more explicitly as a limitation of the significance analysis for NO2 rather than implying a favorable result.

Circularity Check

0 steps flagged · score 0.0 of 10

No load-bearing circularity: the EVT component is a training regularizer fitted to training exceedances and the paper's claims are supported by held-out 2023 forecasts, not by the fit itself.

full rationale

E-STGCN's central derivation is self-contained with respect to its own equations. The GP parameters sigma-hat and xi-hat are fitted by maximum likelihood on training exceedances only, and the POT loss built from them is used as a fixed regularizer in the training objective; the paper's support for E-STGCN over baselines comes from rolling-window forecasts on 2023 data and from MCB and Diebold-Mariano tests, so the held-out evaluation is not statistically forced by the GP fit. No load-bearing step is equivalent to its inputs by construction. The self-citations (Das et al. 2025 and Ray et al. 2023) are peripheral: the MEP discussion they support is explicitly not used for threshold selection because the paper adopts NAAQS thresholds, so these citations are not load-bearing. One non-circular concern is that Eq. (6) evaluates the negative log-likelihood at the raw predicted concentration Z-hat rather than at the exceedance Z-hat minus tau, so the POT term is not exactly the NLL of the fitted GP distribution; this is a model-specification or correctness issue, not a circularity, because it does not make the out-of-sample forecast comparison reduce to the training-time GP fit. The paper also honestly notes limitations, including the poor GP fit for NO2 and the near-equivalence of E-STGCN and modified STGCN for NO2, which further supports treating the extreme-value claims as empirically contingent rather than definitionally forced.

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

The central claim rests on fitted GP parameters, cross-validated loss weights, unreported architecture choices, and assumptions about tail validity and graph construction. There are no invented physical entities, but the model does rely on a policy-chosen threshold that conflicts with standard POT practice.

free parameters (5)
  • GP scale and shape per station (sigma_i, xi_i) = MLE estimates, not tabulated
    Used in Eq. (6) to define the POT loss for each station.
  • POT loss weights beta1 and beta2 = Selected by cross-validation, values not reported
    Controls the trade-off between MSE and GP negative log-likelihood in Eq. (7).
  • Adjacency kernel parameters sigma_tilde^2 and epsilon = Not reported
    Controls the distribution and sparsity of the graph in Eq. (3).
  • LSTM configuration (hidden size, number of layers, lag p) = Not reported
    Defines the capacity of the temporal module and the input window.
  • Conformal window size upsilon = Not reported
    Used in Section 4.6 to compute conformal prediction intervals.
assumptions (4)
  • domain assumption Exceedances above the NAAQS threshold follow a generalized Pareto distribution and are approximately independent.
    Used in Section 3.2.3 and Eq. (6) to construct the POT loss; the Durbin-Watson test checks lag-1 autocorrelation, but no declustering is performed and PM exceedance rates are 61-77%.
  • domain assumption The static Gaussian-kernel graph built from Haversine distances adequately encodes spatial dependence among stations.
    Eqs. (2)-(3) define the adjacency matrix; all spatial information in E-STGCN comes from this static graph.
  • ad hoc to paper A prediction above the threshold can be usefully regularized by the GP likelihood of the fitted extreme distribution.
    This is the design choice in Eq. (7); the paper does not justify why the GP likelihood of a point prediction is a proper scoring rule for threshold exceedances.
  • ad hoc to paper The NAAQS threshold is a valid POT threshold for EVT modeling.
    Section 4.2 explicitly rejects MEP-derived thresholds in favor of NAAQS values, making the EVT approximation an assumption rather than an empirical high-quantile choice.

how reviews work

0 comments
Cite this review

Pith. "Pith review of E-STGCN: Extreme Spatiotemporal Graph Convolutional Networks for Air Quality Forecasting." pith.science (2026). https://pith.science/paper/XVXWWIQ4

@misc{pith2026241112258,
  author       = {Pith},
  title        = {Pith review of: E-STGCN: Extreme Spatiotemporal Graph Convolutional Networks for Air Quality Forecasting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XVXWWIQ4}},
  note         = {Machine review of arXiv:2411.12258}
}
read the original abstract

Modeling and forecasting air quality is crucial for effective air pollution management and protecting public health. Air quality data, characterized by nonlinearity, nonstationarity, and spatiotemporal correlations, often include extreme pollutant levels in severely polluted cities (e.g., Delhi, the capital of India). This is ignored by various geometric deep learning models, such as Spatiotemporal Graph Convolutional Networks (STGCN), which are otherwise effective for spatiotemporal forecasting. This study develops an extreme value theory (EVT) guided modified STGCN model (E-STGCN) for air pollution data to incorporate extreme behavior across pollutant concentrations. E-STGCN combines graph convolutional networks for spatial modeling and EVT-guided long short-term memory units for temporal sequence learning. Along with spatial and temporal components, it incorporates a generalized Pareto distribution to capture the extreme behavior of different air pollutants and embed this information into the learning process. The proposal is then applied to analyze air pollution data of 37 monitoring stations across Delhi, India. The forecasting performance for different test horizons is compared to benchmark forecasters (both temporal and spatiotemporal). It is found that E-STGCN has consistent performance across all seasons. The robustness of our results has also been evaluated empirically. Moreover, combined with conformal prediction, E-STGCN can produce probabilistic prediction intervals.

Figures

Figures reproduced from arXiv: 2411.12258 by the authors.

Figure 1
Figure 1. Extreme Spatiotemporal Graph Convolutional Networks (E-STGCN). Daily air pollution concentration levels from different regions of Delhi, along with the corresponding adjacency matrix, are processed through a Graph Convolu￾tional Network (GCN) and a dense layer to generate spatiotemporal embeddings. To account for extreme values, each sensor’s time series data is modeled using a Generalized Pareto (GP) distribution. … view at source ↗
Figure 2
Figure 2. Dataset split for different forecast evaluation windows [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗
Figure 3
Figure 3. Upper panel: Spatial distribution of the monitoring stations in Delhi and average pollution level of (a) PM [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Block maxima plot for extreme value analysis of (a) PM [PITH_FULL_IMAGE:figures/full_fig_p016_4.png]
Figure 5
Figure 5. Figure 5: Mean excess plot for (a) PM2.5, (b) PM10, and (c) NO2 pollutant concentration in Alipur, Delhi monitoring station. The blue solid line indicates the mean excess level, the red dotted lines represent the 95% confidence interval, and the green solid line is the threshold…
Figure 6
Figure 6. Figure 6: Mean excess plot for NO2 pollutant concentration monitored in (a) Station 4 (Aya Nagar), (b) Station 16 (Lodhi Road IMD), and (c) Station 20 (Najafgarh) of Delhi. The blue solid line indicates the mean excess level, the red dotted lines represent the 95% confidence int…
Figure 7
Figure 7. Figure 7: (a)-(c) Probability density plots of PM2.5, PM10, and NO2 pollutant concentration extremes in Alipur, Delhi moni￾toring station, respectively. All histograms are fitted with the probability density (blue) of the generalized Pareto distribution. 4.3. Forecasting Perform…
Figure 8
Figure 8. Figure 8: MCB Test results for (A) PM2.5, (B) PM10, and (C) NO2 pollutant concentration levels based on RMSE (upper panel) and CRPS (lower panel) metric. In the figure, for example, ‘E-STGCN-3.27’ means that the average rank of the proposed E-STGCN algorithm based on the RMSE er…
Figure 9
Figure 9. Figure 9: DM test results comparing (left) E-STGCN and Modified STGCN, and (right) E-STGCN and NBeats for forecasting [PITH_FULL_IMAGE:figures/full_fig_p027_9.png]
Figure 10
Figure 10. Figure 10: Upper panel presents ground truth (red dots) PM [PITH_FULL_IMAGE:figures/full_fig_p028_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

80 extracted references · 69 canonical work pages

  1. [1]

    , author Chitra, P

    author Abirami, S. , author Chitra, P. , year 2021 . title Regional air quality forecasting using spatiotemporal deep learning . journal Journal of Cleaner Production volume 283 , pages 125341

    work page 2021

  2. [2]

    , author Masseran, N

    author AL-Dhurafi, N.A. , author Masseran, N. , author Zamzuri, Z.H. , author Razali, A.M. , year 2018 . title Modeling unhealthy air pollution index using a peaks-over-threshold method . journal Environmental Engineering Science volume 35 , pages 101--110

    work page 2018

  3. [3]

    , author Karpatne, A

    author Atluri, G. , author Karpatne, A. , author Kumar, V. , year 2018 . title Spatio-temporal data mining: A survey of problems and methods . journal ACM Computing Surveys (CSUR) volume 51 , pages 1--41

    work page 2018

  4. [4]

    , author De Haan, L

    author Balkema, A.A. , author De Haan, L. , year 1974 . title Residual life time at great age . journal The Annals of probability volume 2 , pages 792--804

    work page 1974

  5. [5]

    , author Segerdahl, C.O

    author Benktander, G. , author Segerdahl, C.O. , year 1960 . title On the analytical representation of claim distributions with special reference to excess of loss reinsurance , in: booktitle Transactions of the h\ ternational Congress of Actuaries , pp. pages 1--10

    work page 1960

  6. [6]

    , author Jenkins, G.M

    author Box, G.E. , author Jenkins, G.M. , author Reinsel, G.C. , author Ljung, G.M. , year 1970 . title Time series analysis: forecasting and control . publisher John Wiley & Sons

    work page 1970

  7. [7]

    , author Holgate, S.T

    author Brunekreef, B. , author Holgate, S.T. , year 2002 . title Air pollution and health . journal The lancet volume 360 , pages 1233--1242

    work page 2002

  8. [8]

    , author Schere, K.L

    author Byun, D. , author Schere, K.L. , year 2006 . title Review of the governing equations, computational algorithms, and other components of the models-3 community multiscale air quality (cmaq) modeling system . journal Applied mechanics reviews volume 59 , pages 51--77

    work page 2006

Show all 80 references
  1. [9]

    , year 2012

    author Castillo, E. , year 2012 . title Extreme value theory in engineering . publisher Elsevier

    2012

  2. [10]

    , author Lei, X

    author Chen, J. , author Lei, X. , author Zhang, L. , author Peng, B. , year 2015 . title Using extreme value theory approaches to forecast the probability of outbreak of highly pathogenic influenza in zhejiang, china . journal PloS one volume 10 , pages e0118521

    2015

  3. [11]

    , author Kang, Y

    author Chen, Y. , author Kang, Y. , author Chen, Y. , author Wang, Z. , year 2020 . title Probabilistic forecasting with temporal convolutional neural network . journal Neurocomputing volume 399 , pages 491--501

    2020

  4. [12]

    , author Mwita, P

    author Chukwudum, Q.C. , author Mwita, P. , author Mung’atu, J.K. , year 2020 . title Optimal threshold determination based on the mean excess plot . journal Communications in Statistics-Theory and Methods volume 49 , pages 5948--5963

    2020

  5. [13]

    , author Ord, J

    author Cliff, A. , author Ord, J. , year 1975 . title Model building and the analysis of spatial pattern in human geography . journal Journal of the Royal Statistical Society: Series B (Methodological) volume 37 , pages 297--328

    1975

  6. [14]

    , author Bawa, J

    author Coles, S. , author Bawa, J. , author Trenner, L. , author Dorazio, P. , year 2001 . title An introduction to statistical modeling of extreme values . volume volume 208 . publisher Springer

    2001

  7. [15]

    , author Athulya, R

    author Das, D. , author Athulya, R. , author Chakraborty, T. , author Ray, A. , author Hens, C. , author Dana, S.K. , author Ghosh, D. , author Murukesh, N. , year 2025 . title Pattern change of precipitation extremes in svalbard . journal Scientific Reports volume 15 , pages 8754

    2025

  8. [16]

    , author Mariano, R.S

    author Diebold, F.X. , author Mariano, R.S. , year 2002 . title Comparing predictive accuracy . journal Journal of Business & economic statistics volume 20 , pages 134--144

    2002

  9. [17]

    , author Li, T

    author Du, S. , author Li, T. , author Yang, Y. , author Horng, S.J. , year 2019 . title Deep air quality forecasting using hybrid deep learning framework . journal IEEE Transactions on Knowledge and Data Engineering volume 33 , pages 2412--2424

    2019

  10. [18]

    , author Watson, G.S

    author Durbin, J. , author Watson, G.S. , year 1971 . title Testing for serial correlation in least squares regression. iii . journal Biometrika volume 58 , pages 1--19

    1971

  11. [19]

    , author Burn, D.H

    author Durocher, M. , author Burn, D.H. , author Ashkar, F. , year 2019 . title Comparison of estimation methods for a nonstationary index-flood model in flood frequency analysis using peaks over threshold . journal Water Resources Research volume 55 , pages 9398--9416

    2019

  12. [20]

    , author Kl \"u ppelberg, C

    author Embrechts, P. , author Kl \"u ppelberg, C. , author Mikosch, T. , year 2013 . title Modelling extremal events: for insurance and finance . volume volume 33 . publisher Springer Science & Business Media

    2013

  13. [21]

    , author Rapp, R

    author Faustini, A. , author Rapp, R. , author Forastiere, F. , year 2014 . title Nitrogen dioxide and mortality: review and meta-analysis of long-term studies . journal European Respiratory Journal volume 44 , pages 744--753

    2014

  14. [22]

    , author Tippett, L.H.C

    author Fisher, R.A. , author Tippett, L.H.C. , year 1928 . title Limiting forms of the frequency distribution of the largest or smallest member of a sample , in: booktitle Mathematical proceedings of the Cambridge philosophical society , organization Cambridge University Press...

    1928

  15. [23]

    , year 2000

    author Fletcher, R. , year 2000 . title Practical methods of optimization . publisher John Wiley & Sons

    2000

  16. [24]

    , author Li, W

    author Gao, X. , author Li, W. , year 2021 . title A graph-based lstm model for pm2. 5 forecasting . journal Atmospheric Pollution Research volume 12 , pages 101150

    2021

  17. [25]

    , author Resnick, S

    author Ghosh, S. , author Resnick, S. , year 2010 . title A discussion on mean excess plots . journal Stochastic Processes and their Applications volume 120 , pages 1492--1517

    2010

  18. [26]

    , author Schoenholz, S.S

    author Gilmer, J. , author Schoenholz, S.S. , author Riley, P.F. , author Vinyals, O. , author Dahl, G.E. , year 2017 . title Neural message passing for quantum chemistry , in: booktitle International conference on machine learning , organization PMLR . pp. pages 1263--1272

    2017

  19. [27]

    , author Raftery, A.E

    author Gneiting, T. , author Raftery, A.E. , year 2007 . title Strictly proper scoring rules, prediction, and estimation . journal Journal of the American statistical Association volume 102 , pages 359--378

    2007

  20. [28]

    , author Wolffram, D

    author Gneiting, T. , author Wolffram, D. , author Resin, J. , author Kraus, K. , author Bracher, J. , author Dimitriadis, T. , author Hagenmeyer, V. , author Jordan, A.I. , author Lerch, S. , author Phipps, K. , et al., year 2023 . title Model diagnostics and forecast evaluat...

    2023

  21. [29]

    , year 1993

    author Grimshaw, S.D. , year 1993 . title Computing maximum likelihood estimates for the generalized pareto distribution . journal Technometrics volume 35 , pages 185--191

    1993

  22. [30]

    , year 2018

    author Guinness, J. , year 2018 . title Permutation and grouping methods for sharpening gaussian process approximations . journal Technometrics volume 60 , pages 415--429

    2018

  23. [31]

    , year 1958

    author Gumbel, E.J. , year 1958 . title Statistics of extremes . publisher Columbia university press

    1958

  24. [32]

    , author et al

    author Herzen, J. , author et al. , year 2022 . title Darts: User-friendly modern machine learning for time series . journal Journal of Machine Learning Research volume 23 , pages 1--6

    2022

  25. [33]

    , author Schmidhuber, J

    author Hochreiter, S. , author Schmidhuber, J. , year 1997 . title Long short-term memory . journal Neural computation volume 9 , pages 1735--1780

    1997

  26. [34]

    , year 1980

    author Horowitz, J. , year 1980 . title Extreme values from a nonstationary stochastic process: an application to air quality analysis . journal Technometrics volume 22 , pages 469--478

    1980

  27. [35]

    , year 2018

    author Hyndman, R. , year 2018 . title Forecasting: principles and practice . publisher OTexts

    2018

  28. [36]

    , author Liu, L

    author Ji, J.S. , author Liu, L. , author Zhang, J. , author Kan, H. , author Zhao, B. , author Burkart, K.G. , author Zeng, Y. , year 2022 . title No2 and pm2. 5 air pollution co-exposure and temperature effect modification on pre-mature mortality in advanced age: a longitudi...

    2022

  29. [37]

    , author Koh, H.Y

    author Jin, M. , author Koh, H.Y. , author Wen, Q. , author Zambon, D. , author Alippi, C. , author Webb, G.I. , author King, I. , author Pan, S. , year 2024 . title A survey on graph neural networks for time series: Forecasting, classification, imputation, and anomaly detecti...

    2024

  30. [38]

    , author Chen, B.H

    author Kan, H.D. , author Chen, B.H. , year 2004 . title Statistical distributions of ambient air pollutants in shanghai, china . journal Biomedical and Environmental Sciences volume 17 , pages 366--372

    2004

  31. [39]

    , author Kevrekidis, I.G

    author Karniadakis, G.E. , author Kevrekidis, I.G. , author Lu, L. , author Perdikaris, P. , author Wang, S. , author Yang, L. , year 2021 . title Physics-informed machine learning . journal Nature Reviews Physics volume 3 , pages 422--440

    2021

  32. [40]

    , author Parlange, M.B

    author Katz, R.W. , author Parlange, M.B. , author Naveau, P. , year 2002 . title Statistics of extremes in hydrology . journal Advances in water resources volume 25 , pages 1287--1304

    2002

  33. [41]

    , author Welling, M

    author Kipf, T.N. , author Welling, M. , year 2016 . title Semi-supervised classification with graph convolutional networks . journal arXiv preprint arXiv:1609.02907

    2016 arXiv

  34. [42]

    , author Franses, P.H

    author Koning, A.J. , author Franses, P.H. , author Hibon, M. , author Stekler, H.O. , year 2005 . title The m3 competition: Statistical tests of the results . journal International journal of forecasting volume 21 , pages 397--409

    2005

  35. [43]

    , author Jain, V

    author Kumar, U. , author Jain, V. , year 2010 . title Arima forecasting of ambient air pollutants (o3, no, no2 and co) . journal Stochastic Environmental Research and Risk Assessment volume 24 , pages 751--760

    2010

  36. [44]

    , author Monjardino, J

    author Lei, M.T. , author Monjardino, J. , author Mendes, L. , author Gon c alves, D. , author Ferreira, F. , year 2019 . title Macao air quality forecast using statistical methods . journal Air Quality, Atmosphere & Health volume 12 , pages 1049--1057

    2019

  37. [45]

    , author Evans, J.S

    author Lelieveld, J. , author Evans, J.S. , author Fnais, M. , author Giannadaki, D. , author Pozzer, A. , year 2015 . title The contribution of outdoor air pollution sources to premature mortality on a global scale . journal Nature volume 525 , pages 367--371

    2015

  38. [46]

    , author Peng, L

    author Li, X. , author Peng, L. , author Yao, X. , author Cui, S. , author Hu, Y. , author You, C. , author Chi, T. , year 2017 . title Long short-term memory neural network for air pollutant concentration predictions: Method development and evaluation . journal Environmental ...

    2017

  39. [47]

    , author Raggad, B

    author Marimoutou, V. , author Raggad, B. , author Trabelsi, A. , year 2009 . title Extreme value theory and value at risk: application to oil market . journal Energy Economics volume 31 , pages 519--530

    2009

  40. [48]

    , author Wikuats, C.F.H

    author Martins, L.D. , author Wikuats, C.F.H. , author Capucim, M.N. , author de Almeida, D.S. , author da Costa, S.C. , author Albuquerque, T. , author Carvalho, V.S.B. , author de Freitas, E.D. , author de F \'a tima Andrade, M. , author Martins, J.A. , year 2017 . title Ext...

    2017

  41. [49]

    , author Sun, Y

    author Nag, P. , author Sun, Y. , author Reich, B.J. , year 2023 . title Spatio-temporal deepkriging for interpolation and probabilistic forecasting . journal Spatial Statistics volume 57 , pages 100773

    2023

  42. [50]

    o \"o sli, M. , author Naidoo, R.N. , author K \

    author Olaniyan, T. , author Jeebhay, M. , author R \"o \"o sli, M. , author Naidoo, R.N. , author K \"u nzli, N. , author de Hoogh, K. , author Saucy, A. , author Badpa, M. , author Baatjies, R. , author Parker, B. , et al., year 2020 . title The association between ambient n...

    2020

  43. [51]

    , author Sugiura, K

    author Ong, B.T. , author Sugiura, K. , author Zettsu, K. , year 2016 . title Dynamically pre-trained deep recurrent neural networks using environmental monitoring data for predicting pm 2.5 . journal Neural Computing and Applications volume 27 , pages 1553--1566

    2016

  44. [52]

    , author Carpov, D

    author Oreshkin, B.N. , author Carpov, D. , author Chapados, N. , author Bengio, Y. , year 2019 . title N-beats: Neural basis expansion analysis for interpretable time series forecasting . journal arXiv preprint arXiv:1905.10437

    2019 arXiv

  45. [53]

    , author Brauer, M

    author Pandey, A. , author Brauer, M. , author Cropper, M.L. , author Balakrishnan, K. , author Mathur, P. , author Dey, S. , author Turkgulu, B. , author Kumar, G.A. , author Khare, M. , author Beig, G. , et al., year 2021 . title Health and economic impact of air pollution i...

    2021

  46. [54]

    , author Deutrch, S.J

    author Pfeifer, P.E. , author Deutrch, S.J. , year 1980 . title A three-stage iterative procedure for space-time modeling phillip . journal Technometrics volume 22 , pages 35--47

    1980

  47. [55]

    , year 1975

    author Pickands III, J. , year 1975 . title Statistical inference using extreme order statistics . journal the Annals of Statistics , pages 119--131

    1975

  48. [56]

    , author Perdikaris, P

    author Raissi, M. , author Perdikaris, P. , author Karniadakis, G.E. , year 2019 . title Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations . journal Journal of Computational p...

    2019

  49. [57]

    , author Chakraborty, T

    author Ray, A. , author Chakraborty, T. , author Radhakrishnan, A. , author Hens, C. , author Dana, S.K. , author Ghosh, D. , author Murukesh, N. , year 2023 . title Pattern change of precipitation extremes in bear island . journal arXiv preprint arXiv:2312.04502

    2023 arXiv

  50. [58]

    , author Thomas, M

    author Reiss, R.D. , author Thomas, M. , author Reiss, R. , year 1997 . title Statistical analysis of extreme values . volume volume 2 . publisher Springer

    1997

  51. [59]

    , year 1979

    author Roberts, E. , year 1979 . title Review of statistics of extreme values with applications to air quality data: part ii. applications . journal Journal of the Air Pollution Control Association volume 29 , pages 733--740

    1979

  52. [60]

    , year 2014

    author Rocco, M. , year 2014 . title Extreme value theory in finance: A survey . journal Journal of Economic Surveys volume 28 , pages 82--108

    2014

  53. [61]

    , author Borovkova, S.A

    author Ruchjana, B.N. , author Borovkova, S.A. , author Lopuhaa, H. , author Baskoro, E.T. , author Suprijanto, D. , year 2012 . title Least squares estimation of generalized space time autoregressive (gstar) model and its properties , in: booktitle AIP Conference Proceedings-...

    2012

  54. [62]

    , author Singh, K

    author Saha, A. , author Singh, K. , author Ray, M. , author Rathod, S. , year 2020 . title A hybrid spatio-temporal modelling: an application to space-time rainfall forecasting . journal Theoretical and Applied Climatology volume 142 , pages 1271--1282

    2020

  55. [63]

    , author Flunkert, V

    author Salinas, D. , author Flunkert, V. , author Gasthaus, J. , author Januschowski, T. , year 2020 . title Deepar: Probabilistic forecasting with autoregressive recurrent networks . journal International Journal of Forecasting volume 36 , pages 1181--1191

    2020

  56. [64]

    , author Kumar, G.A

    author Salvi, S. , author Kumar, G.A. , author Dhaliwal, R. , author Paulson, K. , author Agrawal, A. , author Koul, P.A. , author Mahesh, P. , author Nair, S. , author Singh, V. , author Aggarwal, A.N. , et al., year 2018 . title The burden of chronic respiratory diseases and...

    2018

  57. [65]

    , author Panda, A.K

    author Samal, K.K.R. , author Panda, A.K. , author Babu, K.S. , author Das, S.K. , year 2021 . title Multi-output tcn autoencoder for long-term pollution forecasting for multiple sites . journal Urban Climate volume 39 , pages 100943

    2021

  58. [66]

    , author Gori, M

    author Scarselli, F. , author Gori, M. , author Tsoi, A.C. , author Hagenbuchner, M. , author Monfardini, G. , year 2008 . title The graph neural network model . journal IEEE transactions on neural networks volume 20 , pages 61--80

    2008

  59. [67]

    , author Zoras, S

    author Sfetsos, A. , author Zoras, S. , author Bartzis, J.G. , author Triantafyllou, A.G. , year 2006 . title Extreme value modeling of daily pm 10 concentrations in an industrial area . journal Fresenius Environmental Bulletin volume 15 , pages 841--845

    2006

  60. [68]

    , author Thomas, M.L

    author Shaddick, G. , author Thomas, M.L. , author Mudu, P. , author Ruggeri, G. , author Gumy, S. , year 2020 . title Half the world’s population are exposed to increasing air pollution . journal NPJ Climate and Atmospheric Science volume 3 , pages 1--5

    2020

  61. [69]

    , author Khare, M

    author Sharma, P. , author Khare, M. , author Chakrabarti, S. , year 1999 . title Application of extreme value theory for predicting violations of air quality standards for an urban road intersection . journal Transportation Research Part D: Transport and Environment volume 4 ...

    1999

  62. [70]

    , author Lemaitre, M

    author Thomas, M. , author Lemaitre, M. , author Wilson, M.L. , author Viboud, C. , author Yordanov, Y. , author Wackernagel, H. , author Carrat, F. , year 2016 . title Applications of extreme value theory in public health . journal PloS one volume 11 , pages e0159312

    2016

  63. [71]

    , author Fisher, B.E

    author Vardoulakis, S. , author Fisher, B.E. , author Pericleous, K. , author Gonzalez-Flesca, N. , year 2003 . title Modelling air quality in street canyons: a review . journal Atmospheric environment volume 37 , pages 155--182

    2003

  64. [72]

    , year 2017

    author Vaswani, A. , year 2017 . title Attention is all you need . journal Advances in Neural Information Processing Systems

    2017

  65. [73]

    , year 1988

    author Vecchia, A.V. , year 1988 . title Estimation and model identification for continuous spatial processes . journal Journal of the Royal Statistical Society Series B: Statistical Methodology volume 50 , pages 297--312

    1988

  66. [74]

    , author Gammerman, A

    author Vovk, V. , author Gammerman, A. , author Shafer, G. , year 2005 . title Algorithmic learning in a random world . volume volume 29 . publisher Springer

    2005

  67. [75]

    , author Li, J

    author Wang, Z. , author Li, J. , author Wang, Z. , author Yang, W. , author Tang, X. , author Ge, B. , author Yan, P. , author Zhu, L. , author Chen, X. , author Chen, H. , et al., year 2014 . title Modeling study of regional severe hazes over mid-eastern china in january 201...

    2014

  68. [76]

    , author Macready, W.G

    author Wolpert, D.H. , author Macready, W.G. , year 1997 . title No free lunch theorems for optimization . journal IEEE transactions on evolutionary computation volume 1 , pages 67--82

    1997

  69. [77]

    , author Green, B

    author Wu, N. , author Green, B. , author Ben, X. , author O'Banion, S. , year 2020 . title Deep transformer models for time series forecasting: The influenza prevalence case . journal arXiv preprint arXiv:2001.08317

    2020 arXiv

  70. [78]

    , author Yin, H

    author Yu, B. , author Yin, H. , author Zhu, Z. , year 2018 . title Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting , in: booktitle Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence , org...

    2018 doi

  71. [79]

    , author Wang, J

    author Zhou, X. , author Wang, J. , author Wang, J. , author Guan, Q. , year 2024 . title Predicting air quality using a multi-scale spatiotemporal graph attention network . journal Information Sciences volume 680 , pages 121072

    2024

  72. [80]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.