Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper explores the use of sparse high-dimensional techniques to approximate solution operators for partial differential equations (PDEs), leveraging a dimension-incremental framework that integrates product basis expansions with orthogonal matching pursuit (OMP). The authors demonstrate that their method significantly reduces the sample size needed for accurate approximations compared to traditional cubature-based approaches while maintaining competitive accuracy. Numerical evaluations reveal that their approach not only requires fewer PDE solves but also provides interpretable insights into the underlying variables and parameter interactions.
Achieving substantial sample size reduction for PDE solution operators, this method offers a game-changing alternative to traditional cubature techniques without sacrificing accuracy.
We investigate the approximation of solution operators for partial differential equations (PDEs) using sparse high-dimensional techniques. Building on a dimension-incremental framework, we combine product basis expansions with sparse recovery methods, specifically orthogonal matching pursuit (OMP), to substantially reduce the required sample size compared with a previously considered cubature-based approach. We evaluate the resulting method numerically on several examples, comparing it against both cubature-based sparse approximation and Fourier neural operators in terms of accuracy, runtime, and sample size. The experiments show that our approach considerably reduces the number of required PDE solves relative to its predecessor while maintaining competitive accuracy, particularly when the solution admits a sparse representation in the chosen basis. Furthermore, the recovered sparse index sets yield interpretable insights into the relevant variables and parameter interactions.