English

Multiscale analysis of accelerated gradient methods

Optimization and Control 2020-06-16 v3 Numerical Analysis Dynamical Systems Numerical Analysis

Abstract

Accelerated gradient descent iterations are widely used in optimization. It is known that, in the continuous-time limit, these iterations converge to a second-order differential equation which we refer to as the accelerated gradient flow. Using geometric singular perturbation theory, we show that, under certain conditions, the accelerated gradient flow possesses an attracting invariant slow manifold to which the trajectories of the flow converge asymptotically. We obtain a general explicit expression in the form of functional series expansions that approximates the slow manifold to any arbitrary order of accuracy. To the leading order, the accelerated gradient flow reduced to this slow manifold coincides with the usual gradient descent. We illustrate the implications of our results on three examples.

Keywords

Cite

@article{arxiv.1807.11354,
  title  = {Multiscale analysis of accelerated gradient methods},
  author = {Mohammad Farazmand},
  journal= {arXiv preprint arXiv:1807.11354},
  year   = {2020}
}

Comments

Minor revisions; Accepted for Publication in SIAM Journal on Optimization

R2 v1 2026-06-23T03:19:01.231Z