Directional filter combined with active learning for rare failure events

DOI10.1016/J.CMA.2024.117105MaRDI QIDQ6566088zbMATH OpenFDO

Authors Jingwen Song, Yifan Cui, Pengfei Wei, Mohsen Rashki, Weihong Zhang, Michael A. Beer

Publication date 3 July 2024

Published in Computer Methods in Applied Mechanics and Engineering (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/J.CMA.2024.117105

zbMATH Keywords

active learningdirectional filterGaussian process regressionmost probable pointrare failure event

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Fracture and damage (74R99) Stochastic and other probabilistic methods applied to problems in solid mechanics (74S60)

Recommendations

• Active learning method combining Kriging model and multimodal-optimization-based importance sampling for the estimation of small failure probability
• Estimation of small failure probabilities by partially Bayesian active learning line sampling: theory and algorithm
• AK-Gibbs: an active learning kriging model based on Gibbs importance sampling algorithm for small failure probabilities
• Partially Bayesian active learning cubature for structural reliability analysis with extremely small failure probabilities
• Analysis of adaptive directional stratification for the controlled estimation of rare event probab\-ilities

Cites work

• Active-learning-driven surrogate modeling for efficient simulation of parametric nonlinear systems
• Bayesian reinforcement learning reliability analysis
• Estimation of small failure probabilities by partially Bayesian active learning line sampling: theory and algorithm
• Gaussian processes for machine learning.
• General multi-dimensional probability integration by directional simulation
• Structural reliability under combined random load sequences
• The soft Monte Carlo method

Cited in

(3)

• Partially Bayesian active learning cubature for structural reliability analysis with extremely small failure probabilities
• Efficient estimation procedure for failure probability function by an augmented directional sampling
• Transitional active learning of small probabilities

This page was built for publication: Directional filter combined with active learning for rare failure events

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6566088)