A comparison of RESTART implementations

Marnix J.J. Garvels; Dirk P. Kroese

doi:10.1109/WSC.1998.745040

A comparison of RESTART implementations

Marnix J.J. Garvels, Dirk P. Kroese

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

88 Downloads (Pure)

Abstract

The RESTART method is a widely applicable simulation technique for the estimation of rare event probabilities. The method is based on the idea to restart the simulation in certain system states, in order to generate more occurrences of the rare event. One of the main questions for any RESTART implementation is how and when to restart the simulation, in order to achieve the most accurate results for a fixed simulation effort. We investigate and compare, both theoretically and empirically, different implementations of the RESTART method. We find that the original RESTART implementation, in which each path is split into a fixed number of copies, may not be the most efficient one. It is generally better to fix the total simulation effort for each stage of the simulation. Furthermore, given this effort, the best strategy is to restart an equal number of times from each state, rather than to restart each time from a randomly chosen state

Original language	English
Title of host publication	Proceedings of the 1998 Winter Simulation Conference
Editors	D.J. Medeiros, Edward F. Watson, John S. Carson, Mani S. Manivannan
Place of Publication	Enschede
Publisher	IEEE
Pages	601-609
Volume	1
ISBN (Print)	0-7803-5133-9
DOIs	https://doi.org/10.1109/WSC.1998.745040
Publication status	Published - 20 Feb 1998
Event	1998 Winter Simulation Conference - Grand Hyatt, Washington, United States Duration: 13 Dec 1998 → 16 Dec 1998

Conference

Conference	1998 Winter Simulation Conference
Country/Territory	United States
City	Washington
Period	13/12/98 → 16/12/98

Keywords

n/a OA procedure

Access to Document

10.1109/WSC.1998.745040

2 Report

A comparison of RESTART implementations
Garvels, M. J. J. & Kroese, D. P., 1998, Enschede: University of Twente. 24 p. (Memorandum; no. 1443)
Research output: Book/Report › Report › Professional
A comparison of RESTART implementations
Garvels, M. J. J. & Kroese, D. P., 1998, Enschede: Centre for Telematics and Information Technology (CTIT). 24 p. (CTIT technical report series; no. 98-09)
Research output: Book/Report › Report › Other research output

Cite this

@inproceedings{038fd07381b84f60b058c9c95458e7a4,

title = "A comparison of RESTART implementations",

abstract = "The RESTART method is a widely applicable simulation technique for the estimation of rare event probabilities. The method is based on the idea to restart the simulation in certain system states, in order to generate more occurrences of the rare event. One of the main questions for any RESTART implementation is how and when to restart the simulation, in order to achieve the most accurate results for a fixed simulation effort. We investigate and compare, both theoretically and empirically, different implementations of the RESTART method. We find that the original RESTART implementation, in which each path is split into a fixed number of copies, may not be the most efficient one. It is generally better to fix the total simulation effort for each stage of the simulation. Furthermore, given this effort, the best strategy is to restart an equal number of times from each state, rather than to restart each time from a randomly chosen state",

keywords = "n/a OA procedure",

author = "Garvels, {Marnix J.J.} and Kroese, {Dirk P.}",

year = "1998",

month = feb,

day = "20",

doi = "10.1109/WSC.1998.745040",

language = "English",

isbn = "0-7803-5133-9",

volume = "1",

pages = "601--609",

editor = "D.J. Medeiros and Watson, {Edward F.} and Carson, {John S.} and Manivannan, {Mani S.}",

booktitle = "Proceedings of the 1998 Winter Simulation Conference",

publisher = "IEEE",

note = "1998 Winter Simulation Conference ; Conference date: 13-12-1998 Through 16-12-1998",

}

TY - GEN

T1 - A comparison of RESTART implementations

AU - Garvels, Marnix J.J.

AU - Kroese, Dirk P.

PY - 1998/2/20

Y1 - 1998/2/20

N2 - The RESTART method is a widely applicable simulation technique for the estimation of rare event probabilities. The method is based on the idea to restart the simulation in certain system states, in order to generate more occurrences of the rare event. One of the main questions for any RESTART implementation is how and when to restart the simulation, in order to achieve the most accurate results for a fixed simulation effort. We investigate and compare, both theoretically and empirically, different implementations of the RESTART method. We find that the original RESTART implementation, in which each path is split into a fixed number of copies, may not be the most efficient one. It is generally better to fix the total simulation effort for each stage of the simulation. Furthermore, given this effort, the best strategy is to restart an equal number of times from each state, rather than to restart each time from a randomly chosen state

AB - The RESTART method is a widely applicable simulation technique for the estimation of rare event probabilities. The method is based on the idea to restart the simulation in certain system states, in order to generate more occurrences of the rare event. One of the main questions for any RESTART implementation is how and when to restart the simulation, in order to achieve the most accurate results for a fixed simulation effort. We investigate and compare, both theoretically and empirically, different implementations of the RESTART method. We find that the original RESTART implementation, in which each path is split into a fixed number of copies, may not be the most efficient one. It is generally better to fix the total simulation effort for each stage of the simulation. Furthermore, given this effort, the best strategy is to restart an equal number of times from each state, rather than to restart each time from a randomly chosen state

KW - n/a OA procedure

U2 - 10.1109/WSC.1998.745040

DO - 10.1109/WSC.1998.745040

M3 - Conference contribution

SN - 0-7803-5133-9

VL - 1

SP - 601

EP - 609

BT - Proceedings of the 1998 Winter Simulation Conference

A2 - Medeiros, D.J.

A2 - Watson, Edward F.

A2 - Carson, John S.

A2 - Manivannan, Mani S.

PB - IEEE

CY - Enschede

T2 - 1998 Winter Simulation Conference

Y2 - 13 December 1998 through 16 December 1998

ER -