Scalable Approximate Optimal Diagonal Preconditioning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gao, Wenzhi, Qu, Zhaonan, Udell, Madeleine, Ye, Yinyu
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909378504818688
author Gao, Wenzhi
Qu, Zhaonan
Udell, Madeleine
Ye, Yinyu
author_facet Gao, Wenzhi
Qu, Zhaonan
Udell, Madeleine
Ye, Yinyu
contents We consider the problem of finding the optimal diagonal preconditioner for a positive definite matrix. Although this problem has been shown to be solvable and various methods have been proposed, none of the existing approaches are scalable to matrices of large dimension, or when access is limited to black-box matrix-vector products, thereby significantly limiting their practical application. In view of these challenges, we propose practical algorithms applicable to finding approximate optimal diagonal preconditioners of large sparse systems. Our approach is based on the idea of dimension reduction, and combines techniques from semi-definite programming (SDP), random projection, semi-infinite programming (SIP), and column generation. Numerical experiments demonstrate that our method scales to sparse matrices of size greater than $10^7$. Notably, our approach is efficient and implementable using only black-box matrix-vector product operations, making it highly practical for a wide variety of applications.
format Preprint
id arxiv_https___arxiv_org_abs_2312_15594
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Scalable Approximate Optimal Diagonal Preconditioning
Gao, Wenzhi
Qu, Zhaonan
Udell, Madeleine
Ye, Yinyu
Numerical Analysis
Optimization and Control
We consider the problem of finding the optimal diagonal preconditioner for a positive definite matrix. Although this problem has been shown to be solvable and various methods have been proposed, none of the existing approaches are scalable to matrices of large dimension, or when access is limited to black-box matrix-vector products, thereby significantly limiting their practical application. In view of these challenges, we propose practical algorithms applicable to finding approximate optimal diagonal preconditioners of large sparse systems. Our approach is based on the idea of dimension reduction, and combines techniques from semi-definite programming (SDP), random projection, semi-infinite programming (SIP), and column generation. Numerical experiments demonstrate that our method scales to sparse matrices of size greater than $10^7$. Notably, our approach is efficient and implementable using only black-box matrix-vector product operations, making it highly practical for a wide variety of applications.
title Scalable Approximate Optimal Diagonal Preconditioning
topic Numerical Analysis
Optimization and Control
url https://arxiv.org/abs/2312.15594