Lyapunov analyses for first-order methods : Theory, automation, and algorithm design
(2025)- Abstract
- This thesis, comprising four research papers, contributes to the field of systematic and computer-aided analyses and design of first-order methods. The first two papers focus on developing new methodologies, while the remaining two apply these techniques to refine complexity and convergence results, as well as to design new methods. A central theme of the work is the use of Lyapunov-type analyses, a structured proof technique with historical roots in the study of dynamical systems, which is widely used to establish complexity and convergence properties of first-order methods.
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/a006046c-9dc6-446b-b625-f129587c9674
- author
- Upadhyaya, Manu
LU
- supervisor
- opponent
-
- Assistant professor Stellato, Bartolomeo, Princeton University
- organization
- publishing date
- 2025
- type
- Thesis
- publication status
- published
- subject
- pages
- 185 pages
- publisher
- Department of Automatic Control, Lund University
- defense location
- Lecture hall A, building M, Ole Römers väg 1
- defense date
- 2025-09-26 09:15:00
- ISBN
- 978-91-8104-589-5
- 978-91-8104-590-1
- project
- Automatic Lyapunov analysis of optimization algorithms
- language
- English
- LU publication?
- yes
- id
- a006046c-9dc6-446b-b625-f129587c9674
- date added to LUP
- 2025-08-28 15:18:28
- date last changed
- 2025-09-08 16:41:24
@phdthesis{a006046c-9dc6-446b-b625-f129587c9674, abstract = {{This thesis, comprising four research papers, contributes to the field of systematic and computer-aided analyses and design of first-order methods. The first two papers focus on developing new methodologies, while the remaining two apply these techniques to refine complexity and convergence results, as well as to design new methods. A central theme of the work is the use of Lyapunov-type analyses, a structured proof technique with historical roots in the study of dynamical systems, which is widely used to establish complexity and convergence properties of first-order methods.}}, author = {{Upadhyaya, Manu}}, isbn = {{978-91-8104-589-5}}, language = {{eng}}, publisher = {{Department of Automatic Control, Lund University}}, school = {{Lund University}}, title = {{Lyapunov analyses for first-order methods : Theory, automation, and algorithm design}}, url = {{https://lup.lub.lu.se/search/files/226456919/thesis.pdf}}, year = {{2025}}, }