A space decomposition scheme for maximum eigenvalue functions and its applications (Q2407989)
From MaRDI portal
scientific article
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | A space decomposition scheme for maximum eigenvalue functions and its applications |
scientific article |
Statements
A space decomposition scheme for maximum eigenvalue functions and its applications (English)
0 references
9 October 2017
0 references
This paper studies the \(VU\) theory about a special class of eigenvalue function, and the largest eigenvalues with the matrix convex mapping. The authors obtain the first- and second-order derivatives of the \(U\)-Lagrangian. The second-order expansion of the maximum eigenvalue function is derived and the second-order derivatives are explicitly computed. A conceptual algorithm that is proved to have local superlinear convergence is presented. The algorithm is applied to solve the bilinear matrix inequality problem.
0 references
nonsmooth optimization
0 references
eigenvalue optimization
0 references
matrix-convex
0 references
semidefinite programming
0 references
\(\mathcal {VU}\)-decomposition
0 references
\(\mathcal {U}\)-Lagrangian
0 references
smooth manifold
0 references
second-order derivative
0 references
bilinear matrix inequality
0 references
algorithm
0 references
superlinear convergence
0 references
bilinear matrix inequality problem
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
