Constructing confidence intervals for a quantile using batching and sectioning when applying Latin hypercube sampling

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Scopus citations

Abstract

Quantiles are often used in risk evaluation of complex systems. In some situations, as in regulations regarding safety analyses of nuclear power plants, a confidence interval is required for the quantile of the simulation's output variable. In our current paper, we develop methods to construct confidence intervals for quantiles when applying Latin hypercube sampling, a variance reduction technique that extends stratification for sampling in higher dimensions. Our approaches employ the batching and sectioning methods when applying replicated Latin hypercube sampling, with a single Latin hypercube sample in each batch, and samples across batches are independent. We have established the asymptotic validity of the confidence intervals developed in this paper. Moreover, we have proven that quantile estimators from a single Latin hypercube sample and replicated Latin hypercube samples satisfy weak Bahadur representations. An advantage of sectioning over batching is that the sectioning confidence interval typically has better coverage, which we observe in numerical experiments.

Original languageEnglish (US)
Title of host publicationProceedings of the 2014 Winter Simulation Conference, WSC 2014
EditorsAndreas Tolk, Levent Yilmaz, Saikou Y. Diallo, Ilya O. Ryzhov
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages640-651
Number of pages12
ISBN (Electronic)9781479974863
DOIs
StatePublished - Jan 23 2015
Event2014 Winter Simulation Conference, WSC 2014 - Savannah, United States
Duration: Dec 7 2014Dec 10 2014

Publication series

NameProceedings - Winter Simulation Conference
Volume2015-January
ISSN (Print)0891-7736

Other

Other2014 Winter Simulation Conference, WSC 2014
CountryUnited States
CitySavannah
Period12/7/1412/10/14

All Science Journal Classification (ASJC) codes

  • Software
  • Modeling and Simulation
  • Computer Science Applications

Fingerprint Dive into the research topics of 'Constructing confidence intervals for a quantile using batching and sectioning when applying Latin hypercube sampling'. Together they form a unique fingerprint.

Cite this