Semiparametric cross-entropy for rare-event simulation

A.A.N. Ridder, Z. Botev, L. Rojas-Nandayapa

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

The cross entropy is a well-known adaptive importance sampling method which requires estimating an optimal importance sampling distribution within a parametric class. In this paper we analyze an alternative version of the cross entropy, where the importance sampling distribution is selected instead within a general semiparametric class of distributions. We show that the semiparametric cross entropy method delivers efficient estimators in a wide variety of rare-event problems. We illustrate the favourable performance of the method with numerical experiments.
Original languageEnglish
Pages (from-to)633-649
JournalJournal of Applied Probability
Volume53
Issue number3
DOIs
Publication statusPublished - 2016

Fingerprint

Rare Event Simulation
Cross-entropy
Importance Sampling
Sampling Distribution
Cross-entropy Method
Semiparametric Methods
Adaptive Sampling
Efficient Estimator
Rare Events
Sampling Methods
Numerical Experiment
Alternatives
Rare events
Importance sampling
Simulation
Class

Cite this

Ridder, A.A.N. ; Botev, Z. ; Rojas-Nandayapa, L. / Semiparametric cross-entropy for rare-event simulation. In: Journal of Applied Probability. 2016 ; Vol. 53, No. 3. pp. 633-649.
@article{073818751c044aa2bd2bb762aaedac90,
title = "Semiparametric cross-entropy for rare-event simulation",
abstract = "The cross entropy is a well-known adaptive importance sampling method which requires estimating an optimal importance sampling distribution within a parametric class. In this paper we analyze an alternative version of the cross entropy, where the importance sampling distribution is selected instead within a general semiparametric class of distributions. We show that the semiparametric cross entropy method delivers efficient estimators in a wide variety of rare-event problems. We illustrate the favourable performance of the method with numerical experiments.",
author = "A.A.N. Ridder and Z. Botev and L. Rojas-Nandayapa",
year = "2016",
doi = "10.1017/jpr.2016.31",
language = "English",
volume = "53",
pages = "633--649",
journal = "Journal of Applied Probability",
issn = "0021-9002",
publisher = "University of Sheffield",
number = "3",

}

Semiparametric cross-entropy for rare-event simulation. / Ridder, A.A.N.; Botev, Z.; Rojas-Nandayapa, L.

In: Journal of Applied Probability, Vol. 53, No. 3, 2016, p. 633-649.

Research output: Contribution to JournalArticleAcademicpeer-review

TY - JOUR

T1 - Semiparametric cross-entropy for rare-event simulation

AU - Ridder, A.A.N.

AU - Botev, Z.

AU - Rojas-Nandayapa, L.

PY - 2016

Y1 - 2016

N2 - The cross entropy is a well-known adaptive importance sampling method which requires estimating an optimal importance sampling distribution within a parametric class. In this paper we analyze an alternative version of the cross entropy, where the importance sampling distribution is selected instead within a general semiparametric class of distributions. We show that the semiparametric cross entropy method delivers efficient estimators in a wide variety of rare-event problems. We illustrate the favourable performance of the method with numerical experiments.

AB - The cross entropy is a well-known adaptive importance sampling method which requires estimating an optimal importance sampling distribution within a parametric class. In this paper we analyze an alternative version of the cross entropy, where the importance sampling distribution is selected instead within a general semiparametric class of distributions. We show that the semiparametric cross entropy method delivers efficient estimators in a wide variety of rare-event problems. We illustrate the favourable performance of the method with numerical experiments.

U2 - 10.1017/jpr.2016.31

DO - 10.1017/jpr.2016.31

M3 - Article

VL - 53

SP - 633

EP - 649

JO - Journal of Applied Probability

JF - Journal of Applied Probability

SN - 0021-9002

IS - 3

ER -