High-Resolution Inertial Dynamics with Time-Rescaled Gradients for Nonsmooth Convex Optimization

Manh Hung Le, Andrea Simonetto · 2026

Read High-Resolution Inertial Dynamics with Time-Rescaled Gradients for Nonsmooth Convex Optimization by Manh Hung Le, Andrea Simonetto (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-25401.txt.

Description

Le and Simonetto follow a physics-style flow that solves rough convex problems fast. Their system is a high-grade view of Nesterov's acceleration, adapted to objectives with kinks. A time-varying Moreau smoothing softens the kinks, and a rescaled gradient with Hessian damping steers the path. They prove the flow is well posed and build an energy that drains at the right rate. Paths settle weakly onto minimizers under mild conditions. A twin system covers monotone operators, and numeric tests show the knobs at work.

Catalog record

CREATORManh Hung Le, Andrea Simonetto
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE104795
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.25401

Browse the full library · About the Tower of Babel