Coordinatewise descent methods for leading eigenvalue problem


Journal Article

© 2019 Society for Industrial and Applied Mathematics Leading eigenvalue problems for large scale matrices arise in many applications. Coordinatewise descent methods are considered in this work for such problems based on a reformulation of the leading eigenvalue problem as a nonconvex optimization problem. The convergence of several coordinatewise methods is analyzed and compared. Numerical examples of applications to quantum many-body problems demonstrate the efficiency and provide benchmarks of the proposed coordinatewise descent methods.

Full Text

Duke Authors

Cited Authors

  • Yingzhou, LI; Jianfeng, LU; Wang, AZHE

Published Date

  • January 1, 2019

Published In

Volume / Issue

  • 41 / 4

Start / End Page

  • A2681 - A2716

Electronic International Standard Serial Number (EISSN)

  • 1095-7197

International Standard Serial Number (ISSN)

  • 1064-8275

Digital Object Identifier (DOI)

  • 10.1137/18M1202505

Citation Source

  • Scopus