Bounding rare event probabilities in computer experiments

DOI10.1016/J.CSDA.2014.06.023zbMATH Open1506.62012DBLPjournals/csda/AuffrayBM14arXiv1105.0871OpenAlexW2017027134WikidataQ60461469 ScholiaQ60461469MaRDI QIDQ1623705FDOQ1623705

Authors: Yves Auffray, Pierre Barbillon, Jean-Michel Marin 

Publication date: 23 November 2018

Published in: Computational Statistics and Data Analysis (Search for Journal in Brave)

Abstract: We are interested in bounding probabilities of rare events in the context of computer experiments. These rare events depend on the output of a physical model with random input variables. Since the model is only known through an expensive black box function, standard efficient Monte Carlo methods designed for rare events cannot be used. We then propose a strategy to deal with this difficulty based on importance sampling methods. This proposal relies on Kriging metamodeling and is able to achieve sharp upper confidence bounds on the rare event probabilities. The variability due to the Kriging metamodeling step is properly taken into account. The proposed methodology is applied to a toy example and compared to more standard Bayesian bounds. Finally, a challenging real case study is analyzed. It consists of finding an upper bound of the probability that the trajectory of an airborne load will collide with the aircraft that has released it.

Full work available at URL: https://arxiv.org/abs/1105.0871

Recommendations

• Rare event probability estimation in the presence of epistemic uncertainty on input probability distribution parameters
• Determination of sample size in a rare event simulation method
• Error analysis for probabilities of rare events with approximate models
• Bayesian analysis of rare events
• Estimation of rare event probabilities and quantiles applied to an aerospace vehicle fall-back zone analysis

zbMATH Keywords

computer experimentsimportance samplingkrigingBayesian estimatesrare eventsrisk assessment with fighter aircraft

Mathematics Subject Classification ID

Computational methods for problems pertaining to statistics (62-08) Design of statistical experiments (62K99)

Cites Work

• Efficient global optimization of expensive black-box functions
• The design and analysis of computer experiments.
• Design and analysis of computer experiments. With comments and a rejoinder by the authors
• Bayesian calibration of computer models. (With discussion)
• Gaussian processes for machine learning.
• Title not available (Why is that?)
• Sequential design of computer experiments for the estimation of a probability of failure
• Title not available (Why is that?)
• Combining Field Data and Computer Simulations for Calibration and Prediction
• Exploratory designs for computational experiments
• Title not available (Why is that?)
• Adaptive Multilevel Splitting for Rare Event Analysis
• Rare events, splitting, and quasi-Monte Carlo
• A Bayesian calibration approach to the thermal problem
• Uniform Design: Theory and Application
• Cross validation and maximum likelihood estimations of hyper-parameters of Gaussian processes with model misspecification
• Fast simulation of rare events in queueing and reliability models
• Controlled stratification for quantile estimation
• Genealogical particle analysis of rare events
• Orthogonal-maximin Latin hypercube designs
• An efficient methodology for modeling complex computer codes with Gaussian processes
• A Bayesian analysis of the thermal challenge problem
• Maximin design on non hypercube domains and kernel interpolation
• Gaussian process emulation for second-order Monte Carlo simulations
• Estimating Percentiles of Uncertain Computer Code Outputs
• Mechanism-based emulation of dynamic simulation models: concept and application in hydrology
• Adaptive particle techniques and rare event estimation

Cited In (4)

• Bayesian Subset Simulation
• Computing Bounds on Expectations
• Searching for High-Value Rare Events with Uncheatable Grid Computing
• Bayesian analysis of rare events

Uses Software

• DACE
• EGO

This page was built for publication: Bounding rare event probabilities in computer experiments

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1623705)