An Inverse Iteration Method for Eigenvalue Problems with Eigenvector Nonlinearities

  title={An Inverse Iteration Method for Eigenvalue Problems with Eigenvector Nonlinearities},
  author={Elias Jarlebring and Simen Kvaal and Wim Michiels},
Consider a symmetric matrix $A(v)\in\RR^{n\times n}$ depending on a vector $v\in\RR^n$ and satisfying the property $A(\alpha v)=A(v)$ for any $\alpha\in\RR\backslash{0}$. We will here study the problem of finding $(\lambda,v)\in\RR\times \RR^n\backslash\{0\}$ such that $(\lambda,v)$ is an eigenpair of the matrix $A(v)$ and we propose a generalization of inverse iteration for eigenvalue problems with this type of eigenvector nonlinearity. The convergence of the proposed method is studied and… 
Linearizability of eigenvector nonlinearities
A method to linearize a specific class of eigenvalue problems with eigenvector nonlinearities (NEPv), where the non linearities are expressed by scalar functions that are defined by a quotient of linear functions of the eigen vector.
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
The J-method for the Gross-Pitaevskii eigenvalue problem
A local linear rate of convergence is established and global convergence towards an eigenfunction is proved in the case of the Gross-Pitaevskii eigenvalue problem and this quantitative convergence analysis is closely connected to the J-method's unique feature of sensitivity with respect to spectral shifts.
On an eigenvector-dependent nonlinear eigenvalue problem from the perspective of relative perturbation theory
Sufficient conditions on the solvability and solution uniqueness of NEPv are obtained, based on the well-known results from the relative perturbation theory, which is complementary to recent ones in Cai et al. (2018).
Implicit algorithms for eigenvector nonlinearities
Although the modifications of the update equation make the methods implicit, it is shown how corresponding iterates can be computed explicitly, and one of the modifications leads directly to a well-established method (the self-consistent field iteration) whereas the other method is to the authors' knowledge new and has several attractive properties.
Optimal Convergence Rate of Self-Consistent Field Iteration for Solving Eigenvector-dependent Nonlinear Eigenvalue Problems
The effectiveness of the convergence rate estimates is demonstrated numerically for NEPv arising from solving the Kohn-Sham equation in electronic structure calculation and the Gross-Pitaevskii equation in the modeling of Bose-Einstein condensation.
Sobolev Gradient Flow for the Gross-Pitaevskii Eigenvalue Problem: Global Convergence and Computational Efficiency
A new normalized Sobolev gradient flow for the Gross-Pitaevskii eigenvalue problem based on an energy inner product that depends on time through the density of the flow itself and converges to an eigenfunction is proposed.
Homotopy Methods for Eigenvector-Dependent Nonlinear Eigenvalue Problems
Eigenvector-dependent nonlinear eigenvalue problems are considered which arise from the finite difference discretizations of the Gross-Pitaevskii equation and existence and uniqueness of positive and antisymmetric eigenvector for both one and two dimensional cases are proved.
On optimal convergence rates for discrete minimizers of the Gross-Pitaevskii energy in LOD spaces
This paper revisits a two-level discretization based on the Localized Orthogonal Decomposition and proves optimal order error estimates for the L2and H1-error between the exact ground state and discrete minimizers, as well as error Estimates for the ground state energy and the groundstate eigenvalue.
Crank-Nicolson Galerkin approximations to nonlinear Schr\"odinger equations with disorder potentials
This paper analyses the numerical solution of a class of non-linear Schrodinger equations by Galerkin finite elements in space and a mass- and energy conserving variant of the Crank-Nicolson method


For the nonlinear eigenvalue problem $A(\hat \lambda )\hat x = 0$, where $A( \cdot )$ is a matrix-valued operator, residual inverse iteration with shift $\sigma $ is defined by \[ a^{(l + 1)} : =
Inexact Rayleigh Quotient-Type Methods for Eigenvalue Computations
AbstractWe consider the computation of an eigenvalue and corresponding eigenvector of a Hermitian positive definite matrix A ∈ $$\mathbb{C}^{n \times n}$$ , assuming that good approximations of the
Inverse Iteration, Ill-Conditioned Equations and Newton’s Method
Inverse iteration is one of the most widely used algorithms in practical linear algebra but an understanding of its main numerical properties has developed piecemeal over the last thirty years: a
Computing an Eigenvector with Inverse Iteration
It is concluded that the behavior of the residuals in inverse iteration is governed by the departure of the matrix from normality rather than by the conditioning of a Jordan basis or the defectiveness of eigenvalues.
Nonlinear eigenvector algorithms for local optimization in multivariate data analysis
Abstract Many multivariate statistical procedures, such as principal component, canonical correlation, correspondence, and discriminant analysis, are based on the solution of a certain matrix
Convergence factors of Newton methods for nonlinear eigenvalue problems
Abstract Consider a complex sequence { λ k } k = 0 ∞ convergent to λ ∗ ∈ C with order p ∈ N . The convergence factor is typically defined as the fraction c k : = ( λ k + 1 - λ ∗ ) / ( λ k - λ ∗ ) p
Analyzing the convergence factor of residual inverse iteration
We will establish here a formula for the convergence factor of the method called residual inverse iteration, which is a method for nonlinear eigenvalue problems and a generalization of the well-known
An Iteration Method for Solving Nonlinear Eigenvalue Problems
In a recent paper [7] the convergence of an inverse iteration type algorithm for a certain class of nonlinear elliptic eigenvalue problems was discussed. Such algorithms have been used successfully
Convergence of inexact inverse iteration with application to preconditioned iterative solves
In this paper we study inexact inverse iteration for solving the generalised eigenvalue problem Ax=λMx. We show that inexact inverse iteration is a modified Newton method and hence obtain convergence
Global monotone convergence of Newton iteration for a nonlinear eigen-problem
Abstract The nonlinear eigen-problem Ax+F(x)=λx, where A is an n×n irreducible Stieltjes matrix, is considered. Sufficient conditions are given, such that the problem has a unique positive solution