A note on the performance of the gamma kernel estimators at the boundary

Shunpu Zhang

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

The gamma kernel estimator is proposed in Chen [Chen, S.X., 2000. Probability density function estimation using gamma kernels. Annals of the Institute of Statistical Mathematics 52, 471-480] to estimate densities with support [0, ∞). It is shown in his paper that the gamma kernel estimator is non-negative, free of boundary bias, and achieves the optimal rate of convergence for the mean integrated squared error. Numerical results reported in Chen's paper show that, in the boundary region, the gamma kernel estimator even outperforms some widely used boundary corrected density estimators such as the boundary kernel estimator. However, our study finds that the gamma kernel estimator at x = 0 is actually the reflection estimator when the double exponential kernel is used and is only boundary problem free when the estimated density has a shoulder at x = 0 (i.e., the first derivative of the density at x = 0 is zero). For densities not satisfying the shoulder condition, we show that the gamma kernel estimator has a severe boundary problem and its performance is inferior to that of the boundary kernel estimator.

Original languageEnglish (US)
Pages (from-to)548-557
Number of pages10
JournalStatistics and Probability Letters
Volume80
Issue number7-8
DOIs
StatePublished - Jan 1 2010

Fingerprint

Kernel Estimator
Boundary Bias
kernel
Mean Integrated Squared Error
Density Estimates
Optimal Rate of Convergence
Density Estimator
Function Estimation
Density Estimation
Free Boundary Problem
Boundary Problem
Kernel estimator
Probability density function
Non-negative
Estimator
Derivative
Numerical Results
Zero

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

A note on the performance of the gamma kernel estimators at the boundary. / Zhang, Shunpu.

In: Statistics and Probability Letters, Vol. 80, No. 7-8, 01.01.2010, p. 548-557.

Research output: Contribution to journalArticle

@article{41ae3517fa3f4cbc94bb3191d872384f,
title = "A note on the performance of the gamma kernel estimators at the boundary",
abstract = "The gamma kernel estimator is proposed in Chen [Chen, S.X., 2000. Probability density function estimation using gamma kernels. Annals of the Institute of Statistical Mathematics 52, 471-480] to estimate densities with support [0, ∞). It is shown in his paper that the gamma kernel estimator is non-negative, free of boundary bias, and achieves the optimal rate of convergence for the mean integrated squared error. Numerical results reported in Chen's paper show that, in the boundary region, the gamma kernel estimator even outperforms some widely used boundary corrected density estimators such as the boundary kernel estimator. However, our study finds that the gamma kernel estimator at x = 0 is actually the reflection estimator when the double exponential kernel is used and is only boundary problem free when the estimated density has a shoulder at x = 0 (i.e., the first derivative of the density at x = 0 is zero). For densities not satisfying the shoulder condition, we show that the gamma kernel estimator has a severe boundary problem and its performance is inferior to that of the boundary kernel estimator.",
author = "Shunpu Zhang",
year = "2010",
month = "1",
day = "1",
doi = "10.1016/j.spl.2009.12.009",
language = "English (US)",
volume = "80",
pages = "548--557",
journal = "Statistics and Probability Letters",
issn = "0167-7152",
publisher = "Elsevier",
number = "7-8",

}

TY - JOUR

T1 - A note on the performance of the gamma kernel estimators at the boundary

AU - Zhang, Shunpu

PY - 2010/1/1

Y1 - 2010/1/1

N2 - The gamma kernel estimator is proposed in Chen [Chen, S.X., 2000. Probability density function estimation using gamma kernels. Annals of the Institute of Statistical Mathematics 52, 471-480] to estimate densities with support [0, ∞). It is shown in his paper that the gamma kernel estimator is non-negative, free of boundary bias, and achieves the optimal rate of convergence for the mean integrated squared error. Numerical results reported in Chen's paper show that, in the boundary region, the gamma kernel estimator even outperforms some widely used boundary corrected density estimators such as the boundary kernel estimator. However, our study finds that the gamma kernel estimator at x = 0 is actually the reflection estimator when the double exponential kernel is used and is only boundary problem free when the estimated density has a shoulder at x = 0 (i.e., the first derivative of the density at x = 0 is zero). For densities not satisfying the shoulder condition, we show that the gamma kernel estimator has a severe boundary problem and its performance is inferior to that of the boundary kernel estimator.

AB - The gamma kernel estimator is proposed in Chen [Chen, S.X., 2000. Probability density function estimation using gamma kernels. Annals of the Institute of Statistical Mathematics 52, 471-480] to estimate densities with support [0, ∞). It is shown in his paper that the gamma kernel estimator is non-negative, free of boundary bias, and achieves the optimal rate of convergence for the mean integrated squared error. Numerical results reported in Chen's paper show that, in the boundary region, the gamma kernel estimator even outperforms some widely used boundary corrected density estimators such as the boundary kernel estimator. However, our study finds that the gamma kernel estimator at x = 0 is actually the reflection estimator when the double exponential kernel is used and is only boundary problem free when the estimated density has a shoulder at x = 0 (i.e., the first derivative of the density at x = 0 is zero). For densities not satisfying the shoulder condition, we show that the gamma kernel estimator has a severe boundary problem and its performance is inferior to that of the boundary kernel estimator.

UR - http://www.scopus.com/inward/record.url?scp=77049116504&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77049116504&partnerID=8YFLogxK

U2 - 10.1016/j.spl.2009.12.009

DO - 10.1016/j.spl.2009.12.009

M3 - Article

AN - SCOPUS:77049116504

VL - 80

SP - 548

EP - 557

JO - Statistics and Probability Letters

JF - Statistics and Probability Letters

SN - 0167-7152

IS - 7-8

ER -