Resampled regenerative estimators

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

We discuss some estimators for simulations of processes having multiple regenerative sequences. The estimators are obtained by resampling trajectories without and with replacement, which correspond to a type of U-statistic and a type of V-statistic, respectively. The U-statistic estimator turns out to be equivalent to the permuted regenerative estimator, which we previously proposed, but the V-statistic estimator is new. We compare analytically some properties of these estimators and the semiregenerative estimator. We show that when estimating the second moment of a cycle reward, the semiregenerative estimator has positive bias, which is strictly larger than the (positive) bias of the V-statistic estimator. The permuted estimator is unbiased. All of the estimators have the same asymptotic central limit behavior, with reduced asymptotic variance compared to the standard regenerative estimator. Some numerical results are included.

Original languageEnglish (US)
Article number23
JournalACM Transactions on Modeling and Computer Simulation
Volume25
Issue number4
DOIs
StatePublished - May 1 2015

All Science Journal Classification (ASJC) codes

  • Modeling and Simulation
  • Computer Science Applications

Keywords

  • Regenerative method
  • Variance reduction

Fingerprint Dive into the research topics of 'Resampled regenerative estimators'. Together they form a unique fingerprint.

Cite this