Back to search

Cornell University

Some Extensions on the Reach of First-Order Optimization Theory

Abstract

dc:description.abstract

This thesis concerns the foundations of first-order optimization theory. In recent years, these methods have found tremendous success in a wide range of domains due to their incredibly scalable nature. We aim to extend the reach of first-order optimization guarantees, lessening the gap between practice and theory, as well as enabling the design of new algorithms. We show that many of the typical assumptions in the theory literature are not fundamentally needed. For example, analysis in nonsmooth optimization typically relies on Lipschitz continuity, convexity, and carefully chosen stepsize parameters. Chapters 2-4 show that classic methods can be applied and analyzed without relying on these structures. Then Chapters 5-8 consider reformulations of optimization problems that further extend the reach of first-order methods. Constrained optimization can be reformulated to avoid orthogonal projections, generic optimization problems can be reformulated to possess H\"older growth/error bounds, and minimax optimization problems can be smoothed and convexified/concavified. Together these results challenge the limitations on which problems are historically considered tractable for analysis sake.

Degree

thesis:*
Name thesis:degree_name
Ph. D., Operations Research and Information Engineering
Level thesis:degree_level
Doctor of Philosophy
Discipline thesis:degree_discipline
Operations Research and Information Engineering
Grantor
Cornell University
Year dc:date.issued
2021

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Grimmer, Benjamin David
Chair dc:contributor.chair
  • Renegar, James
Committee members dc:contributor.committeemember
  • Davis, Damek Shea
  • Lewis, Adrian S.

Subjects

dc:subject × 4

Rights

dc:rights
Statement dc:rights
  • Attribution-ShareAlike 4.0 International
Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Bibid
15160179
Dc Identifier Other
ProQuest Submission ID: 12557
ProQuest Publication ID: 28545519

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Grimmer, Benjamin David. Some Extensions on the Reach of First-Order Optimization Theory. Doctor of Philosophy thesis, Cornell University, 2021. https://hdl.handle.net/1813/110555