Perturbation analysis of an eigenvector-dependent nonlinear eigenvalue problem with applications

The eigenvector-dependent nonlinear eigenvalue problem arises in many important applications, such as the discretized Kohn–Sham equation in electronic structure calculations and the trace ratio problem in linear discriminant analysis. In this paper, we perform a perturbation analysis for the eigenve...

Full description

Saved in:
Bibliographic Details
Published in:BIT Vol. 60; no. 1; pp. 1 - 29
Main Authors: Cai, Yunfeng, Jia, Zhigang, Bai, Zheng-Jian
Format: Journal Article
Language:English
Published: Dordrecht Springer Netherlands 01.03.2020
Springer Nature B.V
Subjects:
Computational Mathematics and Numerical Analysis
Discriminant analysis
Eigenvalues
Eigenvectors
Electronic structure
Mathematics
Mathematics and Statistics
Numeric Computing
Optimization
Perturbation methods
Upper bounds
47J10
Perturbation analysis
Kohn–Sham equation
65F30
65F15
Trace ratio optimization
Nonlinear eigenvalue problem
15A18
ISSN:0006-3835, 1572-9125
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The eigenvector-dependent nonlinear eigenvalue problem arises in many important applications, such as the discretized Kohn–Sham equation in electronic structure calculations and the trace ratio problem in linear discriminant analysis. In this paper, we perform a perturbation analysis for the eigenvector-dependent nonlinear eigenvalue problem, which gives upper bounds for the distance between the solution to the original nonlinear eigenvalue problem and the solution to the perturbed nonlinear eigenvalue problem. A condition number for the nonlinear eigenvalue problem is introduced, which reveals the factors that affect the sensitivity of the solution. Furthermore, two computable error bounds are given for the nonlinear eigenvalue problem, which can be used to measure the quality of an approximate solution. Numerical results on practical problems, such as the Kohn–Sham equation and the trace ratio optimization, indicate that the proposed upper bounds are sharper than the state-of-the-art bounds.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0006-3835
1572-9125
DOI:10.1007/s10543-019-00765-4