A gradient enhanced ${\ell }_1$-minimization for sparse approximation of polynomial chaos expansions

Abstract

We investigate a gradient-enhanced ${\ell }_1$-minimization for constructing sparse polynomial chaos expansions. In addition to function evaluations, measurements of the function gradient is also included to accelerate the identification of expansion coefficients. By designing appropriate preconditioners to the measurement matrix, we show gradient-enhanced ${\ell }_1$ minimization leads to stable and accurate coefficient recovery. The framework for designing preconditioners is quite general and it applies to recover of functions whose domain is bounded or unbounded. Comparisons between the gradient enhanced approach and the standard ${\ell }_1$-minimization are also presented and numerical examples suggest that the inclusion of derivative information can guarantee sparse recovery at a reduced computational cost.

18 pages,9 figures

References in corpus (3)

• Compressive Sampling of Polynomial Chaos Expansions: Convergence Analysis and Sampling Strategies
• Coherence Motivated Sampling and Convergence Analysis of Least-Squares Polynomial Chaos Regression
• A generalized sampling and preconditioning scheme for sparse approximation of polynomial chaos expansions

Cited by in corpus (5)

• Regression-based sparse polynomial chaos for uncertainty quantification of subsurface flow models
• Sensitivity-enhanced generalized polynomial chaos for efficient uncertainty quantification
• A mixed ${\ell }_1$ regularization approach for sparse simultaneous approximation of parameterized PDEs
• A survey of unsupervised learning methods for high-dimensional uncertainty quantification in black-box-type problems
• Rank adaptive tensor recovery based model reduction for partial differential equations with high-dimensional random inputs