Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Estimating tail probabilities

Technical Report ·
DOI:https://doi.org/10.2172/6568797· OSTI ID:6568797

This paper investigates procedures for univariate nonparametric estimation of tail probabilities. Extrapolated values for tail probabilities beyond the data are also obtained based on the shape of the density in the tail. Several estimators which use exponential weighting are described. These are compared in a Monte Carlo study to nonweighted estimators, to the empirical cdf, to an integrated kernel, to a Fourier series estimate, to a penalized likelihood estimate and a maximum likelihood estimate. Selected weighted estimators are shown to compare favorably to many of these standard estimators for the sampling distributions investigated.

Research Organization:
Pacific Northwest Lab., Richland, WA (USA)
DOE Contract Number:
AC06-76RL01830
OSTI ID:
6568797
Report Number(s):
PNL-4580; ON: DE83005241
Country of Publication:
United States
Language:
English