Polynomial chaos expansions for dependent random variables

DOI10.1016/J.CMA.2019.03.049zbMATH Open1441.65014arXiv1903.09682OpenAlexW2924441742WikidataQ115063497 ScholiaQ115063497MaRDI QIDQ2173605FDOQ2173605

Authors: John D. Jakeman, Fabian Franzelin, Michael S. Eldred, Dirk Plfüger, Akil C. Narayan 

Publication date: 17 April 2020

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

Abstract: Polynomial chaos expansions (PCE) are well-suited to quantifying uncertainty in models parameterized by independent random variables. The assumption of independence leads to simple strategies for evaluating PCE coefficients. In contrast, the application of PCE to models of dependent variables is much more challenging. Three approaches can be used. The first approach uses mapping methods where measure transformations, such as the Nataf and Rosenblatt transformation, can be used to map dependent random variables to independent ones; however we show that this can significantly degrade performance since the Jacobian of the map must be approximated. A second strategy is the class of dominating support methods which build PCE using independent random variables whose distributional support dominates the support of the true dependent joint density; we provide evidence that this approach appears to produce approximations with suboptimal accuracy. A third approach, the novel method proposed here, uses Gram-Schmidt orthogonalization (GSO) to numerically compute orthonormal polynomials for the dependent random variables. This approach has been used successfully when solving differential equations using the intrusive stochastic Galerkin method, and in this paper we use GSO to build PCE using a non-intrusive stochastic collocation method. The stochastic collocation method treats the model as a black box and builds approximations of model output from a set of samples. Building PCE from samples can introduce ill-conditioning which does not plague stochastic Galerkin methods. To mitigate this ill-conditioning we generate weighted Leja sequences, which are nested sample sets, to build accurate polynomial interpolants. We show that our proposed approach produces PCE which are orders of magnitude more accurate than PCE constructed using mapping or dominating support methods.

Full work available at URL: https://arxiv.org/abs/1903.09682

Recommendations

• A polynomial chaos expansion in dependent random variables
• Multivariate polynomial chaos expansions with dependent variables
• Computational aspects for constructing realizations of polynomial chaos in high dimension
• Polynomial chaos expansion of a multimodal random vector
• Practical uncertainty quantification analysis involving statistically dependent random variables

zbMATH Keywords

interpolationuncertainty quantificationquadraturepolynomial chaos expansionLeja sequenceNataf transformation

Mathematics Subject Classification ID

Computational methods for problems pertaining to probability theory (60-08) Numerical solutions to stochastic differential and integral equations (65C30) General theory of stochastic processes (60G07)

Cites Work

• Gaussian processes for machine learning.
• Remarks on a Multivariate Transformation
• Tensor-train decomposition
• Inverse problems: a Bayesian perspective
• Sparse Legendre expansions via \(\ell_1\)-minimization
• A Stochastic Collocation Method for Elliptic Partial Differential Equations with Random Input Data
• The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations
• On the convergence of generalized polynomial chaos expansions
• A Sparse Grid Stochastic Collocation Method for Partial Differential Equations with Random Input Data
• An Anisotropic Sparse Grid Stochastic Collocation Method for Partial Differential Equations with Random Input Data
• Title not available (Why is that?)
• High-Order Collocation Methods for Differential Equations with Random Inputs
• High dimensional polynomial interpolation on sparse grids
• Geometric weakly admissible meshes, discrete least squares approximations and approximate Fekete points
• Adaptive Leja sparse grid constructions for stochastic collocation and high-dimensional approximation
• Title not available (Why is that?)
• Approximation of quantities of interest in stochastic PDEs by the random discrete \(L^2\) projection on polynomial spaces
• Stochastic collocation methods on unstructured grids in high dimensions via interpolation
• On Discrete Least-Squares Projection in Unbounded Domain with Random Evaluations and its Application to Parametric Uncertainty Quantification
• Ridge functions
• Stochastic collocation algorithms using \(\ell_1\)-minimization
• A non-adapted sparse approximation of PDEs with stochastic inputs
• Computing multivariate Fekete and Leja points by numerical linear algebra
• Computing approximate Fekete points by QR factorizations of Vandermonde matrices
• Non-intrusive low-rank separated approximation of high-dimensional stochastic models
• Monge-Ampère measures for convex bodies and Bernstein-Markov type inequalities
• Adaptive Smolyak Pseudospectral Approximations
• Sparse pseudospectral approximation method
• Adaptive construction of surrogates for the Bayesian solution of inverse problems
• Sur certaines suites liées aux ensembles plans et leur application à la représentation conforme
• Compressive sampling of polynomial chaos expansions: convergence analysis and sampling strategies
• A generalized sampling and preconditioning scheme for sparse approximation of polynomial chaos expansions
• Time-dependent generalized polynomial chaos
• Enhancing \(\ell_1\)-minimization estimates of polynomial chaos expansions using basis selection
• Numerical approach for quantification of epistemic uncertainty
• Fast Leja points
• Newton interpolation at Leja points
• Subsampled Gauss quadrature nodes for estimating polynomial chaos expansions
• A least-squares method for sparse low rank approximation of multivariate functions
• A flexible numerical approach for quantification of epistemic uncertainty
• Generation and application of multivariate polynomial quadrature rules
• Gradient-based optimization for regression in the functional tensor-train format
• Numerical integration in multiple dimensions with designed quadrature
• Optimal weighted least-squares methods
• A Christoffel function weighted least squares algorithm for collocation approximations
• Measure transformation and efficient quadrature in reduced-dimensional stochastic modeling of coupled problems
• Goal-oriented adaptive surrogate construction for stochastic inversion
• Efficient uncertainty propagation for network multiphysics systems
• Effectively subsampled quadratures for least squares polynomial approximations
• Weighted approximate Fekete points: sampling for least-squares polynomial approximation

Cited In (31)

• Past, current and future trends and challenges in non-deterministic fracture mechanics: a review
• A polynomial chaos expansion in dependent random variables
• Multivariate polynomial chaos expansions with dependent variables
• Parametrization of Random Vectors in Polynomial Chaos Expansions via Optimal Transportation
• Non-intrusive reduced-order modeling for uncertainty quantification of space-time-dependent parameterized problems
• Practical uncertainty quantification analysis involving statistically dependent random variables
• Polynomial chaos expansion approach to interest rate models
• Global sensitivity analysis on numerical solver parameters of particle-in-cell models in particle accelerator systems
• On spectral approximations with nonstandard weight functions and their implementations to generalized chaos expansions
• Towards optimal sampling for learning sparse approximation in high dimensions
• High-order spectral method of density estimation for stochastic differential equation driven by multivariate Gaussian random variables
• Design under uncertainty employing stochastic expansion methods
• Generalized polynomial chaos expansion by reanalysis using static condensation based on substructuring
• An Adaptive Sampling and Domain Learning Strategy for Multivariate Function Approximation on Unknown Domains
• Emulation of stochastic simulators using generalized lambda models
• Is it worthwhile considering orthogonality in generalised polynomial chaos expansions applied to solving stochastic models?
• Bayesian inversion using adaptive polynomial chaos kriging within subset simulation
• Uncertainty quantification for random Hamiltonian systems by using polynomial expansions and geometric integrators
• An hp‐adaptive multi‐element stochastic collocation method for surrogate modeling with information re‐use
• Polynomial chaos expansion of a multimodal random vector
• Stochastic dynamic analysis of composite plates in thermal environments using nonlinear autoregressive model with exogenous input in polynomial chaos expansion surrogate
• Polynomial chaos expansions on principal geodesic Grassmannian submanifolds for surrogate modeling and uncertainty quantification
• Polynomial Approximation for a Class of Physical Random Variables
• A Stieltjes Algorithm for Generating Multivariate Orthogonal Polynomials
• Computational aspects for constructing realizations of polynomial chaos in high dimension
• Multifidelity uncertainty quantification with models based on dissimilar parameters
• A Comparative Study of Polynomial-Type Chaos Expansions for Indicator Functions
• Data-driven projection pursuit adaptation of polynomial chaos expansions for dependent high-dimensional parameters
• Variance-based adaptive sequential sampling for polynomial chaos expansion
• Fast sparse grid operations using the unidirectional principle: a generalized and unified framework
• Cholesky-based experimental design for Gaussian process and kernel-based emulation and calibration

This page was built for publication: Polynomial chaos expansions for dependent random variables

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2173605)