Large deviations principle for the Adaptive Multilevel Splitting Algorithm in an idealized setting

MaRDI QIDQ3465405zbMATH OpenFDO

Authors Charles-Edouard Bréhier

Publication date 21 January 2016

Full work available at URL https://arxiv.org/abs/1502.06780, http://alea.impa.br/articles/v12/12-27.pdf

zbMATH Keywords

large deviationsMonte Carlo simulationmultilevel splittingrare events

Mathematics Subject Classification ID

Large deviations (60F10) Order statistics; empirical distribution functions (62G30) Monte Carlo methods (65C05) Stochastic particle methods (65C35)

Abstract: The Adaptive Multilevel Splitting (AMS) algorithm is a powerful and versatile method for the simulation of rare events. It is based on an interacting (via a mutation-selection procedure) system of replicas, and depends on two integer parameters: n $in$ N * the size of the system and the number k $in$ {1, . . . , n -- 1} of the replicas that are eliminated and resampled at each iteration. In an idealized setting, we analyze the performance of this algorithm in terms of a Large Deviations Principle when n goes to infinity, for the estimation of the (small) probability P(X extgreater{} a) where a is a given threshold and X is real-valued random variable. The proof uses the technique introduced in [BLR15]: in order to study the log-Laplace transform, we rely on an auxiliary functional equation. Such Large Deviations Principle results are potentially useful to study the algorithm beyond the idealized setting, in particular to compute rare transitions probabilities for complex high-dimensional stochastic processes.

Recommendations

• A large deviations perspective on the efficiency of multilevel splitting
• Analysis of adaptive multilevel splitting algorithms in an idealized case
• On the asymptotic normality of adaptive multilevel splitting
• Central limit theorem for adaptive multilevel splitting estimators in an idealized setting
• Adaptive multilevel Monte Carlo methods for stochastic variational inequalities
• Unbiasedness of some generalized adaptive multilevel splitting algorithms
• Fluctuation analysis of adaptive multilevel splitting
• General multilevel adaptations for stochastic approximation algorithms. II: CLTs
• Multidimensional stochastic approximation
• On multivariate adaptive approximation

Cited in

(10)

• Application of adaptive multilevel splitting to high-dimensional dynamical systems
• Analysis of adaptive multilevel splitting algorithms in an idealized case
• Central limit theorem for adaptive multilevel splitting estimators in an idealized setting
• On the asymptotic normality of adaptive multilevel splitting
• On a new class of score functions to estimate tail probabilities of some stochastic processes with adaptive multilevel splitting
• Computing return times or return periods with rare event algorithms
• Coupling rare event algorithms with data-based learned committor functions using the analogue Markov chain
• Unbiasedness of some generalized adaptive multilevel splitting algorithms
• Adaptive multilevel splitting: historical perspective and recent results
• Adaptive multilevel splitting in molecular dynamics simulations

This page was built for publication: Large deviations principle for the Adaptive Multilevel Splitting Algorithm in an idealized setting

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3465405)