Density estimation in uncertainty propagation problems using a surrogate model

DOI10.1137/18M1205959zbMATH Open1483.65024OpenAlexW3005256403MaRDI QIDQ4960992FDOQ4960992

Authors: A. Ditkowski, Gadi Fibich, Amir Sagiv 

Publication date: 24 April 2020

Published in: SIAM/ASA Journal on Uncertainty Quantification (Search for Journal in Brave)

Abstract: The effect of uncertainties and noise on a quantity of interest (model output) is often better described by its probability density function (PDF) than by its moments. Although density estimation is a common task, the adequacy of approximation methods (surrogate models) for density estimation has not been analyzed before in the uncertainty-quantification (UQ) literature. In this paper, we first show that standard surrogate models (such as generalized polynomial chaos), which are highly accurate for moment estimation, might completely fail to approximate the PDF, even for one-dimensional noise. This is because density estimation requires that the surrogate model accurately approximates the gradient of the quantity of interest, and not just the quantity of interest itself. Hence, we develop a novel spline-based algorithm for density-estimation whose convergence rate in $L^q$ is polynomial in the sampling resolution. This convergence rate is better than that of standard statistical density-estimation methods (such as histograms and kernel density estimators) at dimensions $1leqdleqfrac52m$, where $m$ is the spline order. Furthermore, we obtain the convergence rate for density estimation with any surrogate model that approximates the quantity of interest and its gradient in $L^{infty}$. Finally, we demonstrate our algorithm for problems in nonlinear optics and fluid dynamics.

Full work available at URL: https://arxiv.org/abs/1803.10991

Recommendations

• Estimation of a density using an improved surrogate model
• Propagation of uncertainties using improved surrogate models
• Piecewise polynomial approximation of probability density functions with application to uncertainty quantification for stochastic PDEs
• A surrogate accelerated multicanonical Monte Carlo method for uncertainty quantification
• Estimation of an improved surrogate model in uncertainty quantification by neural networks

zbMATH Keywords

density estimationuncertainty quantificationprobability density functionnonlinear dynamicssurrogate modelspline

Mathematics Subject Classification ID

Density estimation (62G07) Numerical computation using splines (65D07) Lasers, masers, optical bistability, nonlinear optics (78A60) Applications to the sciences (65Z05)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• $L^\infty $-Multivariate Approximation Theory
• A practical guide to splines
• A primer on radial basis functions with applications to the geosciences
• An adaptive minimum spanning tree multielement method for uncertainty quantification of smooth and discontinuous responses
• An adaptive multi-element generalized polynomial chaos method for stochastic differential equations
• Approximation theory and approximation practice
• Asymptotics in statistics: some basic concepts
• Asynchronous time integration for polynomial chaos expansion of uncertain periodic dynamics
• Error estimates and condition numbers for radial basis function interpolation
• Handbook of uncertainty quantification. In 2 volumes
• High-Order Collocation Methods for Differential Equations with Random Inputs
• Interacting nonlinear wave envelopes and rogue wave formation in deep water
• Introduction to nonparametric estimation
• Monotone Piecewise Cubic Interpolation
• Multiple steady states for characteristic initial value problems
• Numerical Challenges in the Use of Polynomial Chaos Representations for Stochastic Processes
• Numerical methods for stochastic computations. A spectral method approach.
• On Information and Sufficiency
• On cubic spline functions that vanish at all knots
• On spectral approximations with nonstandard weight functions and their implementations to generalized chaos expansions
• On the Convergence of Some Cubic Spline Interpolation Schemes
• On the convergence rates of Legendre approximation
• Optimal error bounds for cubic spline interpolation
• POD-Galerkin modeling and sparse-grid collocation for a natural convection problem with stochastic boundary conditions
• Sparse grid collocation schemes for stochastic natural convection problems
• Sparse pseudospectral approximation method
• Spectral Methods for Time-Dependent Problems
• Spectral Methods for Uncertainty Quantification
• The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations
• The nonlinear Schrödinger equation. Singular solutions and optical collapse
• The stochastic finite element method: past, present and future
• Uncertainty Quantification and Polynomial Chaos Techniques in Computational Fluid Dynamics
• Uncertainty analysis for the steady-state flows in a dual throat nozzle
• Uncertainty propagation using Wiener-Haar expansions
• Uncertainty quantification of geochemical and mechanical compaction in layered sedimentary basins
• Which Cubic Spline should One Use?

Cited In (15)

• Propagation of uncertainties using improved surrogate models
• Nonparametric density estimation and bandwidth selection with B-spline bases: a novel Galerkin method
• Spectral convergence of probability densities for forward problems in uncertainty quantification
• New high-order numerical methods for hyperbolic systems of nonlinear PDEs with uncertainties
• High-order spectral method of density estimation for stochastic differential equation driven by multivariate Gaussian random variables
• The Wasserstein distances between pushed-forward measures with applications to uncertainty quantification
• A surrogate accelerated multicanonical Monte Carlo method for uncertainty quantification
• Hyperbolic balance laws: interplay between scales and randomness. Abstracts from the workshop held February 25 -- March 1, 2024
• Estimation of an improved surrogate model in uncertainty quantification by neural networks
• Estimation of a density using an improved surrogate model
• A review of uncertainty quantification for density estimation
• Piecewise polynomial approximation of probability density functions with application to uncertainty quantification for stochastic PDEs
• Surrogate-Based Ensemble Grouping Strategies for Embedded Sampling-Based Uncertainty Quantification
• Quasi-interpolation for multivariate density estimation on bounded domain
• Convergence of probability densities using approximate models for forward and inverse problems in uncertainty quantification

Uses Software

• pchip

This page was built for publication: Density estimation in uncertainty propagation problems using a surrogate model

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4960992)