Fast parameter estimation of Generalized Extreme Value distribution using Neural Networks

The heavy-tailed behavior of the generalized extreme-value distribution makes it a popular choice for modeling extreme events such as floods, droughts, heatwaves, wildfires, etc. However, estimating the distribution's parameters using conventional maximum likelihood methods can be computationally intensive, even for moderate-sized datasets. To overcome this limitation, we propose a computationally efficient, likelihood-free estimation method utilizing a neural network. Through an extensive simulation study, we demonstrate that the proposed neural network-based method provides Generalized Extreme Value (GEV) distribution parameter estimates with comparable accuracy to the conventional maximum likelihood method but with a significant computational speedup. To account for estimation uncertainty, we utilize parametric bootstrapping, which is inherent in the trained network. Finally, we apply this method to 1000-year annual maximum temperature data from the Community Climate System Model version 3 (CCSM3) across North America for three atmospheric concentrations: 289 ppm $\mathrm{CO}_2$ (pre-industrial), 700 ppm $\mathrm{CO}_2$ (future conditions), and 1400 ppm $\mathrm{CO}_2$, and compare the results with those obtained using the maximum likelihood approach.

[1]  B. Shaby,et al.  Modeling Extremal Streamflow using Deep Learning Approximations and a Flexible Spatial Process , 2022, 2208.03344.

[2]  J. Jalbert,et al.  A flexible extended generalized Pareto distribution for tail estimation , 2022, Environmetrics.

[3]  C. Wikle,et al.  Statistical Deep Learning for Spatial and Spatio-Temporal Data , 2022, Annual Review of Statistics and Its Application.

[4]  Raphael Huser,et al.  A flexible Bayesian hierarchical modeling framework for spatially dependent peaks-over-threshold data , 2021, 2112.09530.

[5]  G. Meehl,et al.  Decadal climate variability in the tropical Pacific: Characteristics, causes, predictability, and prospects , 2021, Science.

[6]  Johann Rudi,et al.  Neural networks for parameter estimation in intractable models , 2021, Comput. Stat. Data Anal..

[7]  H. Rue,et al.  Practical strategies for GEV-based regression models for extremes , 2021, 2106.13110.

[8]  Douglas W. Nychka,et al.  Fast covariance parameter estimation of spatial Gaussian process models using neural networks , 2020, Stat.

[9]  Yanan Fan,et al.  Overview of ABC , 2018, Handbook of Approximate Bayesian Computation.

[10]  H. Rue,et al.  INLA goes extreme: Bayesian tail regression for the estimation of high spatio-temporal quantiles , 2018, Extremes.

[11]  Dirk Roos,et al.  Deep Gaussian Covariance Network , 2017, ArXiv.

[12]  M. Creel.  Neural nets for indirect inference , 2017 .

[13]  Johan Segers,et al.  On the maximum likelihood estimator for the Generalized Extreme-Value distribution , 2016, 1601.05702.

[14]  Elisabeth J. Moyer,et al.  Estimating changes in temperature extremes from millennial scale climate simulations using generalized extreme value (GEV) distributions , 2015, 1512.08775.

[15]  Bai Jiang,et al.  Learning Summary Statistic for Approximate Bayesian Computation via Deep Neural Network , 2015, 1510.02175.

[16]  Anthony C. Davison,et al.  Statistics of Extremes , 2015, International Encyclopedia of Statistical Science.

[17]  Scott A. Sisson,et al.  Modelling extremes using approximate Bayesian Computation , 2014, 1411.1451.

[18]  Michael Creel,et al.  Indirect likelihood inference (revised) , 2013 .

[19]  Richard L. Smith,et al.  Approximate Bayesian computing for spatial extremes , 2011, Comput. Stat. Data Anal..

[20]  Richard W. Katz,et al.  Statistics of extremes in climate change , 2010 .

[21]  John A. Nelder,et al.  Nelder-Mead algorithm , 2009, Scholarpedia.

[22]  T. Ouarda,et al.  Joint Bayesian model selection and parameter estimation of the generalized extreme value model with covariates using birth‐death Markov chain Monte Carlo , 2009 .

[23]  M. Blum.  Approximate Bayesian Computation: A Nonparametric Perspective , 2009, 0904.0635.

[24]  D. Nychka,et al.  Bayesian Spatial Modeling of Extreme Precipitation Return Levels , 2007 .

[25]  W. Collins,et al.  The Community Climate System Model Version 3 (CCSM3) , 2006 .

[26]  W. Adger,et al.  Social Capital, Collective Action, and Adaptation to Climate Change , 2003 .

[27]  Eric P. Smith,et al.  An Introduction to Statistical Modeling of Extreme Values , 2002, Technometrics.

[28]  J. Stedinger,et al.  Generalized maximum‐likelihood generalized extreme‐value quantile estimators for hydrologic data , 2000 .

[29]  Stuart G. Coles,et al.  Spatial Regression Models for Extremes , 1999 .

[30]  G. Geoffrey Booth,et al.  The behavior of extreme values in Germany's stock index futures: An application to intradaily margin setting , 1998 .

[31]  R.J. Cohen,et al.  Linear and nonlinear ARMA model parameter estimation using an artificial neural network , 1997, IEEE Transactions on Biomedical Engineering.

[32]  J. Angus.  Extreme Value Theory in Engineering , 1990 .

[33]  J. R. Wallis,et al.  Estimation of the generalized extreme-value distribution by the method of probability-weighted moments , 1985 .

[34]  Richard L. Smith.  Maximum likelihood estimation in a class of nonregular cases , 1985 .

[35]  A. Jenkinson.  The frequency distribution of the annual maximum (or minimum) values of meteorological elements , 1955 .

[36]  R. Fisher,et al.  Limiting forms of the frequency distribution of the largest or smallest member of a sample , 1928, Mathematical Proceedings of the Cambridge Philosophical Society.

[37]  A. Zammit‐Mangion,et al.  Fast Optimal Estimation with Intractable Models using Permutation-Invariant Neural Networks , 2022 .

[38]  R. Huser,et al.  A unifying partially-interpretable framework for neural network-based extreme quantile regression , 2022, ArXiv.

[39]  J. Kyselý.  NOTES AND CORRESPONDENCE A Cautionary Note on the Use of Nonparametric Bootstrap for Estimating Uncertainties in Extreme-Value Models , 2008 .

[40]  Richard L. Smith.  EXTREME VALUE THEORY , 2008 .

[41]  M. Kenward,et al.  An Introduction to the Bootstrap , 2007 .

[42]  Daniel T. Kaplan,et al.  Introduction to Statistical Modeling , 2005 .

[43]  Vijay P. Singh,et al.  Generalized Extreme Value Distribution , 1998 .

[44]  Richard L. Smith.  Weibull regression models for reliability data , 1991 .