Mathematics > Numerical Analysis
[Submitted on 28 Apr 2016 (v1), revised 28 Nov 2016 (this version, v2), latest version 20 Jun 2017 (v4)]
Title:Sparse Quadrature for High-Dimensional Integration with Gaussian Measure
View PDFAbstract:In this work we analyze the dimension-independent convergence property of a sparse quadrature for numerical integration of functions of high-dimensional parameters with Gaussian measure. We propose both a priori construction and goal-oriented a posteriori construction of the sparse quadrature by a sum of tensorization of univariate quadrature in a carefully selected multi-index set. Under certain assumption of the exactness and the boundedness of the univariate quadrature rules as well as the regularity of the parametric function with respect to the parameters, we obtain the convergence rate $O(N^{-s})$, where $N$ is the number of indices (or quadrature points) and $s$ is independent of the parameter dimension. Examples of nonlinear parametric function and parametric partial differential equations are provided to illustrate and verify the regularity assumption. Numerical experiments are performed to demonstrate the dimension-independent convergence of the sparse quadrature errors. The convergence is shown to be much faster than that of Monte Carlo quadrature errors for the test functions with sufficient regularity with respect to the parameters.
Submission history
From: Peng Chen [view email][v1] Thu, 28 Apr 2016 15:29:38 UTC (22,260 KB)
[v2] Mon, 28 Nov 2016 16:57:38 UTC (2,436 KB)
[v3] Tue, 7 Mar 2017 19:45:57 UTC (2,440 KB)
[v4] Tue, 20 Jun 2017 22:21:56 UTC (2,442 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.