Importance resampling using Chi-Square Tilting

Hani M. Samawi, Ahmad S Bany Mustafa, Mohammad S. Ahmed

Research output: Contribution to journalArticle

Abstract

Significant reduction in the value of the number of bootstrap resamples may be achieved when exponential tilting is used to estimate a tail probability or a quantile, associated with a small probability, say p. However, exponential tilting is based on the assumption that the bootstrap distribution is approximately normal. In this paper, chisquare tilting is introduced when the bootstrap distribution of the statistics of interest is assumed to have a skewed distribution. In the case of small value of the new tilting reduces the asymptotic variance by at least a factor of 6, implies a large saving in bootstrap resamples. Calculations show that the variance reduction is very close to the asymptotic one even for sample sizes as small as 15. Some asymptotic results are given. Computer simulation as well as real data is used to illustrate the computation for the lower bootstrap confidence limit of the bootstrap sample variance. Although, bootstrap confidence intervals for a variance are known to perform very badly in practice even when sampling from a normal distribution, see Schenker (1987), our simulation indicates that our method perform reasonably well.

Original languageEnglish
Pages (from-to)181-198
Number of pages18
JournalMetron
Volume60
Issue number3-4
Publication statusPublished - 2002

Fingerprint

Chi-square
Tilting
Resampling
Bootstrap
Exponential Tilting
Bootstrap Confidence Intervals
Sample variance
Skewed Distribution
Confidence Limits
Variance Reduction
Tail Probability
Asymptotic Variance
Quantile
Gaussian distribution
Sample Size
Computer Simulation
Statistics
Imply
Estimate
Simulation

Keywords

  • Bootstrap
  • Chi-square Tilting
  • Confidence limits
  • Exponential Tilting
  • Importance resample
  • Sample variance

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

Samawi, H. M., Mustafa, A. S. B., & Ahmed, M. S. (2002). Importance resampling using Chi-Square Tilting. Metron, 60(3-4), 181-198.

Importance resampling using Chi-Square Tilting. / Samawi, Hani M.; Mustafa, Ahmad S Bany; Ahmed, Mohammad S.

In: Metron, Vol. 60, No. 3-4, 2002, p. 181-198.

Research output: Contribution to journalArticle

Samawi, HM, Mustafa, ASB & Ahmed, MS 2002, 'Importance resampling using Chi-Square Tilting', Metron, vol. 60, no. 3-4, pp. 181-198.
Samawi HM, Mustafa ASB, Ahmed MS. Importance resampling using Chi-Square Tilting. Metron. 2002;60(3-4):181-198.
Samawi, Hani M. ; Mustafa, Ahmad S Bany ; Ahmed, Mohammad S. / Importance resampling using Chi-Square Tilting. In: Metron. 2002 ; Vol. 60, No. 3-4. pp. 181-198.
@article{30183005edf6447bb91d0d1b94701320,
title = "Importance resampling using Chi-Square Tilting",
abstract = "Significant reduction in the value of the number of bootstrap resamples may be achieved when exponential tilting is used to estimate a tail probability or a quantile, associated with a small probability, say p. However, exponential tilting is based on the assumption that the bootstrap distribution is approximately normal. In this paper, chisquare tilting is introduced when the bootstrap distribution of the statistics of interest is assumed to have a skewed distribution. In the case of small value of the new tilting reduces the asymptotic variance by at least a factor of 6, implies a large saving in bootstrap resamples. Calculations show that the variance reduction is very close to the asymptotic one even for sample sizes as small as 15. Some asymptotic results are given. Computer simulation as well as real data is used to illustrate the computation for the lower bootstrap confidence limit of the bootstrap sample variance. Although, bootstrap confidence intervals for a variance are known to perform very badly in practice even when sampling from a normal distribution, see Schenker (1987), our simulation indicates that our method perform reasonably well.",
keywords = "Bootstrap, Chi-square Tilting, Confidence limits, Exponential Tilting, Importance resample, Sample variance",
author = "Samawi, {Hani M.} and Mustafa, {Ahmad S Bany} and Ahmed, {Mohammad S.}",
year = "2002",
language = "English",
volume = "60",
pages = "181--198",
journal = "Metron",
issn = "0026-1424",
publisher = "Universita di Roma {"}La Sapienza{"}",
number = "3-4",

}

TY - JOUR

T1 - Importance resampling using Chi-Square Tilting

AU - Samawi, Hani M.

AU - Mustafa, Ahmad S Bany

AU - Ahmed, Mohammad S.

PY - 2002

Y1 - 2002

N2 - Significant reduction in the value of the number of bootstrap resamples may be achieved when exponential tilting is used to estimate a tail probability or a quantile, associated with a small probability, say p. However, exponential tilting is based on the assumption that the bootstrap distribution is approximately normal. In this paper, chisquare tilting is introduced when the bootstrap distribution of the statistics of interest is assumed to have a skewed distribution. In the case of small value of the new tilting reduces the asymptotic variance by at least a factor of 6, implies a large saving in bootstrap resamples. Calculations show that the variance reduction is very close to the asymptotic one even for sample sizes as small as 15. Some asymptotic results are given. Computer simulation as well as real data is used to illustrate the computation for the lower bootstrap confidence limit of the bootstrap sample variance. Although, bootstrap confidence intervals for a variance are known to perform very badly in practice even when sampling from a normal distribution, see Schenker (1987), our simulation indicates that our method perform reasonably well.

AB - Significant reduction in the value of the number of bootstrap resamples may be achieved when exponential tilting is used to estimate a tail probability or a quantile, associated with a small probability, say p. However, exponential tilting is based on the assumption that the bootstrap distribution is approximately normal. In this paper, chisquare tilting is introduced when the bootstrap distribution of the statistics of interest is assumed to have a skewed distribution. In the case of small value of the new tilting reduces the asymptotic variance by at least a factor of 6, implies a large saving in bootstrap resamples. Calculations show that the variance reduction is very close to the asymptotic one even for sample sizes as small as 15. Some asymptotic results are given. Computer simulation as well as real data is used to illustrate the computation for the lower bootstrap confidence limit of the bootstrap sample variance. Although, bootstrap confidence intervals for a variance are known to perform very badly in practice even when sampling from a normal distribution, see Schenker (1987), our simulation indicates that our method perform reasonably well.

KW - Bootstrap

KW - Chi-square Tilting

KW - Confidence limits

KW - Exponential Tilting

KW - Importance resample

KW - Sample variance

UR - http://www.scopus.com/inward/record.url?scp=84891468601&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84891468601&partnerID=8YFLogxK

M3 - Article

VL - 60

SP - 181

EP - 198

JO - Metron

JF - Metron

SN - 0026-1424

IS - 3-4

ER -