Efficient implementations of the Multivariate Decomposition Method for approximating infinite-variate integrals

TL;DR

This paper presents efficient algorithms for the Multivariate Decomposition Method (MDM) to approximate infinite-dimensional integrals, optimizing computation by exploiting structure and avoiding redundant evaluations, with demonstrated numerical success.

Contribution

The paper introduces computationally efficient implementations of MDM for infinite-variate integrals, leveraging structure to reduce evaluations and improve performance.

Findings

Efficient MDM algorithms reduce computational cost.

Numerical results confirm effectiveness of the proposed methods.

Structured implementation avoids repeated function evaluations.

Abstract

In this paper we focus on efficient implementations of the Multivariate Decomposition Method (MDM) for approximating integrals of $\infty$-variate functions. Such $\infty$-variate integrals occur for example as expectations in uncertainty quantification. Starting with the anchored decomposition $f = \sum_{\mathfrak{u}\subset\mathbb{N}} f_\mathfrak{u}$, where the sum is over all finite subsets of $\mathbb{N}$ and each $f_\mathfrak{u}$ depends only on the variables $x_j$ with $j\in\mathfrak{u}$, 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 $f_\mathfrak{u}$ term-by-term using Smolyak or (randomized) quasi-Monte Carlo (QMC) quadratures. The anchored decomposition allows us to compute $f_\mathfrak{u}$ 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 $f_\mathfrak{u}$ 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.