论文信息 - Using extreme value theory to estimate large percentiles

Using extreme value theory to estimate large percentiles

Weissman (1978) suggested percentile estimators based on the joint limiting distribution of the k largest order statistics. The present work identifies situations where Weissman's estimators are a significant improvement over the usual sample percentile estimators and gives practical advice on how to use these new estimators effectively. In particular, large reductions in mean squared error can be made when the tails of the distributions are approximately exponential and p ≥ .95.

Dennis D. Boos | D. Boos

[1] H. A. David,et al. Order Statistics (2nd ed). , 1981 .

[2] Leo Breiman,et al. New Methods for Estimating Tail Probabilities and Extreme Value Disributions. , 1979 .

[3] I. Weissman. Estimation of Parameters and Large Quantiles Based on the k Largest Observations , 1978 .

[4] M. Stephens. Tests of fit for the logistic distribution based on the empirical distribution function , 1979 .

[5] A. Azzalini. A note on the estimation of a distribution function and quantiles by a kernel method , 1981 .

[6] A. M. Gross. A Monte Carlo Swindle for Estimators of Location , 1973 .

[7] Herbert Solomon,et al. Approximations to Density Functions Using Pearson Curves , 1978 .

[8] D. Boos. On Weissman's method of estimating large percentiles , 1981 .

[9] Linus Schrage,et al. A More Portable Fortran Random Number Generator , 1979, TOMS.

[10] D. Dickey. Histograms, Percentiles, and Moments , 1981 .

[11] D. Boos. Minimum anderson-darling estimation , 1982 .

[12] J. D. T. Oliveira,et al. The Asymptotic Theory of Extreme Order Statistics , 1979 .

[13] Further Development of New Methods for Estimating Tail Probabilities and Extreme Value Distributions. , 1981 .

[14] George Marsaglia,et al. Improvements on Fast Methods for Generating Normal Random Variables , 1976, Inf. Process. Lett..

[15] I. Weissman. Estimation of tail parameters under type i censoring , 1980 .