English

An efficient method for block low-rank approximations for kernel matrix systems

Numerical Analysis 2018-11-13 v1

Abstract

In the iterative solution of dense linear systems from boundary integral equations or systems involving kernel matrices, the main challenges are the expensive matrix-vector multiplication and the storage cost which are usually tackled by hierarchical matrix techniques such as H\mathcal{H} and H2\mathcal{H}^2 matrices. However, hierarchical matrices also have a high construction cost that is dominated by the low-rank approximations of the sub-blocks of the kernel matrix. In this paper, an efficient method is proposed to give a low-rank approximation of the kernel matrix block K(X0,Y0)K(X_0, Y_0) in the form of an interpolative decomposition (ID) for a kernel function K(x,y)K(x,y) and two properly located point sets X0,Y0X_0, Y_0. The proposed method combines the ID using strong rank-revealing QR (sRRQR), which is purely algebraic, with analytic kernel information to reduce the construction cost of a rank-rr approximation from O(rX0Y0)O(r|X_0||Y_0|), for ID using sRRQR alone, to O(rX0)O(r|X_0|) which is not related to Y0|Y_0|. Numerical experiments show that H2\mathcal{H}^2 matrix construction with the proposed algorithm only requires a computational cost linear in the matrix dimension.

Keywords

Cite

@article{arxiv.1811.04134,
  title  = {An efficient method for block low-rank approximations for kernel matrix systems},
  author = {Xin Xing and Edmond Chow},
  journal= {arXiv preprint arXiv:1811.04134},
  year   = {2018}
}
R2 v1 2026-06-23T05:11:03.688Z