Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
6928774 | Journal of Computational Physics | 2018 | 16 Pages |
Abstract
We investigate a gradient enhanced â1-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 â1 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 â1-minimization are also presented and numerical examples suggest that the inclusion of derivative information can guarantee sparse recovery at a reduced computational cost.
Related Topics
Physical Sciences and Engineering
Computer Science
Computer Science Applications
Authors
Ling Guo, Akil Narayan, Tao Zhou,