The application of Bayesian interpolation in Monte Carlo simulations

M. Rajabalinejad, P.H.A.J.M. van Gelder, N. van Erp

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

1 Citation (Scopus)
139 Downloads (Pure)

Abstract

To reduce the cost of Monte Carlo (MC) simulations for time-consuming processes (like Finite Elements), a Bayesian interpolation method is coupled with the Monte Carlo technique. It is, therefore, possible to reduce the number of realizations in MC by interpolation. Besides, there is a possibility of thought about priors. In other words, this study tries to speed up the Monte Carlo process by taking into the account the prior knowledge about the problem and reduce the number of simulations. Moreover, the information of previous simulations aids to judge accuracy of the prediction in every step. As a result, a narrower confidence interval comes with a higher number of simulations. This paper shows the general methodology, algorithm, and result of the suggested approach in the form of a numerical example
Original languageEnglish
Title of host publicationSafety, reliability and risk analysis
Subtitle of host publicationtheory, methods and applications
EditorsSebastian Martorell, C. Guedes Soares, Julie Barnett
PublisherCRC Press
Pages717-723
Number of pages3357
ISBN (Electronic)9781482266481
ISBN (Print)9780415485135
Publication statusPublished - 2009
Externally publishedYes
EventJoint ESREL 2008 & 17th Society for Risk Analysis, SRA-Europe Conference 2008
- Valencia, Spain
Duration: 22 Sep 200825 Sep 2008

Conference

ConferenceJoint ESREL 2008 & 17th Society for Risk Analysis, SRA-Europe Conference 2008
Abbreviated titleSRA-E
CountrySpain
CityValencia
Period22/09/0825/09/08

Keywords

  • IR-89100

Fingerprint Dive into the research topics of 'The application of Bayesian interpolation in Monte Carlo simulations'. Together they form a unique fingerprint.

Cite this