Synthesis of the Transient Behaviour of Optimization Algorithms for Strongly Convex Objective Functions
Abstract
We study the transient behavior of first-order optimization methods for unconstrained minimization of smooth and strongly convex objective functions. These methods are modelled as linear systems over finite-time intervals embedded in feedback with the gradient nonlinearity. By characterizing the function class through discrete-time Zames-Falb integral quadratic constraints, we employ a robust ℋ2 control framework to obtain convex semidefinite programs certifying tight worst-case bounds on the terminal distance to the optimizer. Building on the same formulation, we derive a convex synthesis procedure for a transient-optimal method that achieves the optimal finite-horizon bounds and guarantees error reduction over any prescribed number of steps. The framework naturally extends to delayed gradient measurements, enabling both transient performance analysis and synthesis of optimal algorithms under feedback delays.
Related
- A sequential quadratic programming method for optimization with expectation objective functions, deterministic inequality constraints and robust subproblems
- Decision-focused predictions via pessimistic bilevel optimization: complexity and algorithms
- A connectedness result for convex polynomial multi-objective optimization without boundedness
- Representing the Non-dominated Set of Multi-objective Optimization Problems with Polyhedral Ordering Cones
- MPFA: A Pareto Front Approximation Method for Riemannian Bi-objective Optimization
BibTeX
@misc{abdelsamed2026synthesis,
title = {{Synthesis of the Transient Behaviour of Optimization Algorithms for Strongly Convex Objective Functions}},
author = {Adel Abdelsamed and Dennis Gramlich and Christian Ebenbauer and Carsten W. Scherer},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.23206},
eprint = {2610.23206},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.23206},
}
RIS
TY - UNPB
TI - Synthesis of the Transient Behaviour of Optimization Algorithms for Strongly Convex Objective Functions
AU - Adel Abdelsamed
AU - Dennis Gramlich
AU - Christian Ebenbauer
AU - Carsten W. Scherer
PY - 2026
DA - 2026/10/16
PB - arXiv
DO - 10.48550/arxiv.2610.23206
UR - https://arxiv.org/abs/2610.23206
AB - We study the transient behavior of first-order optimization methods for unconstrained minimization of smooth and strongly convex objective functions. These methods are modelled as linear systems over finite-time intervals embedded in feedback with the gradient nonlinearity. By characterizing the function class through discrete-time Zames-Falb integral quadratic constraints, we employ a robust ℋ_2 control framework to obtain convex semidefinite programs certifying tight worst-case bounds on the terminal distance to the optimizer. Building on the same formulation, we derive a convex synthesis procedure for a transient-optimal method that achieves the optimal finite-horizon bounds and guarantees error reduction over any prescribed number of steps. The framework naturally extends to delayed gradient measurements, enabling both transient performance analysis and synthesis of optimal algorithms under feedback delays.
ER -