English

A Control-Theoretic Perspective on Optimal High-Order Optimization

Optimization and Control 2026-01-21 v5 Computational Complexity Data Structures and Algorithms

Abstract

We provide a control-theoretic perspective on optimal tensor algorithms for minimizing a convex function in a finite-dimensional Euclidean space. Given a function Φ:RdR\Phi: \mathbb{R}^d \rightarrow \mathbb{R} that is convex and twice continuously differentiable, we study a closed-loop control system that is governed by the operators Φ\nabla \Phi and 2Φ\nabla^2 \Phi together with a feedback control law λ()\lambda(\cdot) satisfying the algebraic equation (λ(t))pΦ(x(t))p1=θ(\lambda(t))^p\|\nabla\Phi(x(t))\|^{p-1} = \theta for some θ(0,1)\theta \in (0, 1). Our first contribution is to prove the existence and uniqueness of a local solution to this system via the Banach fixed-point theorem. We present a simple yet nontrivial Lyapunov function that allows us to establish the existence and uniqueness of a global solution under certain regularity conditions and analyze the convergence properties of trajectories. The rate of convergence is O(1/t(3p+1)/2)O(1/t^{(3p+1)/2}) in terms of objective function gap and O(1/t3p)O(1/t^{3p}) in terms of squared gradient norm. Our second contribution is to provide two algorithmic frameworks obtained from discretization of our continuous-time system, one of which generalizes the large-step A-HPE framework and the other of which leads to a new optimal pp-th order tensor algorithm. While our discrete-time analysis can be seen as a simplification and generalization of~\citet{Monteiro-2013-Accelerated}, it is largely motivated by the aforementioned continuous-time analysis, demonstrating the fundamental role that the feedback control plays in optimal acceleration and the clear advantage that the continuous-time perspective brings to algorithmic design. A highlight of our analysis is that we show that all of the pp-th order optimal tensor algorithms that we discuss minimize the squared gradient norm at a rate of O(k3p)O(k^{-3p}), which complements the recent analysis.

Keywords

Cite

@article{arxiv.1912.07168,
  title  = {A Control-Theoretic Perspective on Optimal High-Order Optimization},
  author = {Tianyi Lin and Michael. I. Jordan},
  journal= {arXiv preprint arXiv:1912.07168},
  year   = {2026}
}

Comments

Accepted by Mathematical Programming Series A; 45 pages

R2 v1 2026-06-23T12:46:38.497Z