EURO-Online login
- New to EURO? Create an account
- I forgot my username and/or my password.
- Help with cookies
(important for IE8 users)
4315. Non-Euclidean Gradient Methods for matrix completion
Invited abstract in session WA-41: Convex optimization algorithms, stream Nonsmooth Optimization.
Wednesday, 8:30-10:00Room: 97 (building: 306)
Authors (first author is the speaker)
| 1. | Susan Ghaderi
|
| ESAT, KU Leuven | |
| 2. | Yves Moreau
|
| KU Leuven | |
| 3. | Masoud Ahookhosh
|
| Department of Mathematics, University of Antwerp |
Abstract
This study develops and analyzes non-Euclidean gradient methods for minimizing relatively smooth nonconvex functions using Bregman distances. We explore both constant and dynamic step-size strategies, examining their convergence and complexity, particularly under relative strong convexity. We demonstrate the methods' effectiveness in matrix factorization, matrix completion, and image inpainting on real datasets, validating our theoretical findings.
Keywords
- Non-smooth Optimization
- Convex Optimization
- Algorithms
Status: accepted
Back to the list of papers