An official website of the United States government
Abstract 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 and so forth. 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 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 across North America for three atmospheric concentrations: 289 ppm (pre‐industrial), 700 ppm (future conditions), and 1400 ppm , and compare the results with those obtained using the maximum likelihood approach.
more »
« less
Uniqueness and global optimality of the maximum likelihood estimator for the generalized extreme value distribution
Summary The three-parameter generalized extreme value distribution arises from classical univariate extreme value theory, and is in common use for analysing the far tail of observed phenomena, yet important asymptotic properties of likelihood-based estimation under this standard model have not been established. In this paper we prove that the maximum likelihood estimator is global and unique. An interesting secondary result entails the uniform consistency of a class of limit relations in a tight neighbourhood of the true shape parameter.
more »
« less
- Award ID(s):
- 2001433
- PAR ID:
- 10376256
- Date Published:
- Journal Name:
- Biometrika
- Volume:
- 109
- Issue:
- 3
- ISSN:
- 0006-3444
- Page Range / eLocation ID:
- 853 to 864
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Massive calving events result in significant instantaneous ice loss from Antarctica. The rarity and stochastic nature of these extreme events makes it difficult to understand their physical drivers, temporal trends, and future likelihood. To address this challenge, we turn to extreme value theory to investigate past trends in annual maxima iceberg area and assess the likelihood of high‐magnitude calving events. We use 47 years of iceberg size from satellite observations. Our analysis reveals no upward trend in the surface area of the largest annual iceberg over this time frame. This finding suggests that extreme calving events such as the recent 2017 Larsen C iceberg, A68, are statistically unexceptional and that extreme calving events are not necessarily a consequence of climate change. Nevertheless, it is statistically possible for Antarctica to experience a calving event up to several times greater than any in the observational record.more » « less
-
Prescreening is a methodology where forensic examiners select samples similar to given trace evidence to represent the background population. This background evidence helps assign a value of evidence using a likelihood ratio or Bayes factor. A key advantage of prescreening is its ability to mitigate effects from subpopulation structures within the alternative source population by isolating the relevant subpopulation. This paper examines the impact of prescreening before assigning evidence value. Extensive simulations with synthetic and real data, including trace element and fingerprint score examples, were conducted. The findings indicate that prescreening can provide an accurate evidence value in cases of subpopulation structures but may also yield more extreme or dampened evidence values within specific subpopulations. The study suggests that prescreening is beneficial for presenting evidence relative to the subpopulation of interest, provided the prescreening method and level are transparently reported alongside the evidence value.more » « less
-
In computational mechanics, multiple models are often present to describe a physical system. While Bayesian model selection is a helpful tool to compare these models using measurement data, it requires the computationally expensive estimation of a multidimensional integral — known as the marginal likelihood or as the model evidence (i.e., the probability of observing the measured data given the model) — over the multidimensional parameter domain. This study presents efficient approaches for estimating this marginal likelihood by transforming it into a one-dimensional integral that is subsequently evaluated using a quadrature rule at multiple adaptively-chosen iso-likelihood contour levels. Three different algorithms are proposed to estimate the probability mass at each adapted likelihood level using samples from importance sampling, stratified sampling, and Markov chain Monte Carlo (MCMC) sampling, respectively. The proposed approach is illustrated — with comparisons to Monte Carlo, nested, and MultiNest sampling — through four numerical examples. The first, an elementary example, shows the accuracies of the three proposed algorithms when the exact value of the marginal likelihood is known. The second example uses an 11-story building subjected to an earthquake excitation with an uncertain hysteretic base isolation layer with two models to describe the isolation layer behavior. The third example considers flow past a cylinder when the inlet velocity is uncertain. Based on the these examples, the method with stratified sampling is by far the most accurate and efficient method for complex model behavior in low dimension, particularly considering that this method can be implemented to exploit parallel computation. In the fourth example, the proposed approach is applied to heat conduction in an inhomogeneous plate with uncertain thermal conductivity modeled through a 100 degree-of-freedom Karhunen–Loève expansion. The results indicate that MultiNest cannot efficiently handle the high-dimensional parameter space, whereas the proposed MCMC-based method more accurately and efficiently explores the parameter space. The marginal likelihood results for the last three examples — when compared with the results obtained from standard Monte Carlo sampling, nested sampling, and MultiNest algorithm — show good agreement.more » « less
-
Abstract High-fidelity simulators that connect theoretical models with observations are indispensable tools in many sciences. If the likelihood is known, inference can proceed using standard techniques. However, when the likelihood is intractable or unknown, a simulator makes it possible to infer the parameters of a theoretical model directly from real and simulated observations when coupled with machine learning. We introduce an extension of the recently proposed likelihood-free frequentist inference (LF2I) approach that makes it possible to construct confidence sets with thep-value function and to use the same function to check the coverage explicitly at any given parameter point. LikeLF2I, this extension yields provably valid confidence sets in parameter inference problems for which a high-fidelity simulator is available. The utility of our algorithm is illustrated by applying it to three pedagogically interesting examples: the first is from cosmology, the second from high-energy physics and astronomy, both with tractable likelihoods, while the third, with an intractable likelihood, is from epidemiology33Code to reproduce all of our results is available onhttps://github.com/AliAlkadhim/ALFFI..more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1093/biomet/asab043
