Threshold selection in univariate extreme value analysis

DOI10.1007/S10687-021-00405-7zbMATH Open1482.62054arXiv1903.02517OpenAlexW3153806905MaRDI QIDQ826008FDOQ826008

Authors: Laura Fee Schneider, Andrea Krajina, Tatyana Krivobokova 

Publication date: 18 December 2021

Published in: Extremes (Search for Journal in Brave)

Abstract: Threshold selection plays a key role for various aspects of statistical inference of rare events. Most classical approaches tackling this problem for heavy-tailed distributions crucially depend on tuning parameters or critical values to be chosen by the practitioner. To simplify the use of automated, data-driven threshold selection methods, we introduce two new procedures not requiring the manual choice of any parameters. The first method measures the deviation of the log-spacings from the exponential distribution and achieves good performance in simulations for estimating high quantiles. The second approach smoothly estimates the asymptotic mean square error of the Hill estimator and performs consistently well over a wide range of distributions. The methods are compared to existing procedures in an extensive simulation study and applied to a dataset of financial losses, where the underlying extreme value index is assumed to vary over time. This application strongly emphasizes the importance of solid automated threshold selection.

Full work available at URL: https://arxiv.org/abs/1903.02517

Recommendations

• Threshold selection in extreme value analysis
• A diagnostic for selecting the threshold in extreme value analysis
• Threshold selection and trimming in extremes
• Threshold selection for extremes under a semiparametric model
• Threshold selection for multivariate heavy-tailed data
• scientific article; zbMATH DE number 3917511
• scientific article; zbMATH DE number 3905657
• Adaptive Threshold Estimation via Extreme Value Theory

zbMATH Keywords

bias estimationHill estimatortuning parameter selectionpower lawspeaks-over-threshold approach

Mathematics Subject Classification ID

Asymptotic properties of parametric estimators (62F12) Nonparametric estimation (62G05) Statistics of extreme values; tail inference (62G32) Applications of statistics to actuarial sciences and financial mathematics (62P05)

Cites Work

• Automated threshold selection for extreme value analysis via ordered goodness-of-fit tests with adjustment for false discovery rate
• Power-law distributions in empirical data
• A review of extreme value threshold estimation and uncertainty quantification
• Extreme value theory. An introduction.
• Statistical inference using extreme order statistics
• Improved threshold diagnostic plots for extreme value analyses
• A simple general approach to inference about the tail of a distribution
• On exponential representations of log-spacings of extreme order statistics
• Semi-parametric estimation of the second order parameter in statistics of extremes
• Nonparametric regression with correlated errors.
• A new class of semi-parametric estimators of the second order parameter.
• Selecting the optimal sample fraction in univariate extreme value estimation
• How to make a Hill plot.
• A diagnostic for selecting the threshold in extreme value analysis
• Comparison of tail index estimators
• On optimising the estimation of high quantiles of a probability distribution
• A semi-parametric estimator of a shape second-order parameter
• Linking Pareto-tail kernel goodness-of-fit statistics with tail index at optimal threshold and second order estimation
• Using a bootstrap method to choose the sample fraction in tail index estimation
• Estimating the stable index \(\alpha\) in order to measure tail thickness: a critique
• Title not available (Why is that?)
• Smoothing the Hill Estimator
• Title not available (Why is that?)
• Alternatives to a semi-parametric estimator of parameters of rare events -- the jackknife methodology
• The qq-estimator and heavy tails
• On the estimation of extreme tail probabilities
• A Note on Penalized Spline Smoothing With Correlated Errors
• Statistics of Heteroscedastic Extremes
• Reiss and Thomas' automatic selection of the number of extremes
• Bandwidth selection for kernel regression with correlated errors

Cited In (14)

• Threshold selection for extremes under a semiparametric model
• Threshold selection in extreme value analysis
• Unimodal thresholding
• Threshold selection and trimming in extremes
• A review of extreme value threshold estimation and uncertainty quantification
• Title not available (Why is that?)
• A computational approach to confidence intervals and testing for generalized Pareto index using the Greenwood statistic
• Multiple thresholds in extremal parameter estimation
• A diagnostic for selecting the threshold in extreme value analysis
• Improving the traditional Hill estimator in threshold selection in extreme value theory
• Reduced bias estimation of the shape parameter of the log-logistic distribution
• Automated threshold selection for extreme value analysis via ordered goodness-of-fit tests with adjustment for false discovery rate
• The risk function of the goodness-of-fit tests for tail models
• Threshold selection for regional peaks-over-threshold data

This page was built for publication: Threshold selection in univariate extreme value analysis

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q826008)