Title
Efficient implementations of the Multivariate Decomposition Method for approximating infinite-variate integrals.
Abstract
In this paper we focus on efficient implementations of the multivariate decomposition method (MDM) for approximating integrals of infinity-variate functions. Such infinity-variate integrals occur, for example, as expectations in uncertainty quantification. Starting with the anchored decomposition f = Sigma(u subset of N) f(u), where the sum is over all finite subsets of N and each f(u) depends only on the variables x(j) with j is an element of u, our MDM algorithm approximates the integral of f by first truncating the sum to some "active set" and then approximating the integral of the remaining functions f(u) term-byterm using Smolyak or (randomized) quasi-Monte Carlo quadratures. The anchored decomposition allows us to compute f(u) explicitly by function evaluations of f. Given the specification of the active set and theoretically derived parameters of the quadrature rules, we exploit structures in both the formula for computing f(u) and the quadrature rules to develop computationally efficient strategies to implement the MDM in various scenarios. In particular, we avoid repeated function evaluations at the same point. We provide numerical results for a test function to demonstrate the effectiveness of the algorithm.
Year
DOI
Venue
2017
10.1137/17M1161890
SIAM JOURNAL ON SCIENTIFIC COMPUTING
Keywords
Field
DocType
quadrature,infinite-variate integral,multivariate decomposition method,quasi-Monte Carlo,Smolyak's method,sparse grids
Discrete mathematics,Random variate,Uncertainty quantification,Mathematical analysis,Multivariate statistics,Test functions for optimization,Decomposition method (constraint satisfaction),Quadrature (mathematics),Mathematics
Journal
Volume
Issue
ISSN
40
5
1064-8275
Citations 
PageRank 
References 
0
0.34
5
Authors
4
Name
Order
Citations
PageRank
Alexander D. Gilbert100.68
Frances Y. Kuo247945.19
Dirk Nuyens316817.97
Grzegorz W. Wasilkowski4527167.51