Abstract
In this paper, we propose the descent method with new inexact line-search for unconstrained optimization problems on Riemannian manifolds. The global convergence of the proposed method is established under some appropriate assumptions. We further analyze some convergence rates, namely R-linear convergence rate, superlinear convergence rate and quadratic convergence rate, of the proposed descent method.
| Original language | English |
|---|---|
| Pages (from-to) | 830-854 |
| Number of pages | 25 |
| Journal | Journal of Optimization Theory and Applications |
| Volume | 180 |
| Issue number | 3 |
| DOIs | |
| State | Published - 15 Mar 2019 |
Bibliographical note
Publisher Copyright:© 2018, Springer Science+Business Media, LLC, part of Springer Nature.
Keywords
- Convergence rate
- Descent method
- New inexact line-search
- Riemannian manifolds
ASJC Scopus subject areas
- Management Science and Operations Research
- Control and Optimization
- Applied Mathematics