Uniform-in-time weak propagation of chaos for consensus-based optimization

Erhan Bayraktar, Ibrahim Ekren, Hongyi Zhou · 2025

Read Uniform-in-time weak propagation of chaos for consensus-based optimization by Erhan Bayraktar, Ibrahim Ekren, Hongyi Zhou (2025) 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-2502-00582.txt.

Description

Erhan Bayraktar, Ibrahim Ekren, and Hongyi Zhou prove a math result in a 2025 paper on a search method called consensus-based optimization. In that method, many points move and pull toward a shared weighted center, in hopes of settling at the best point. The question is how well a finite swarm of points tracks the ideal limit over time.

The authors show that the weak error stays of order one over the number of points, with no growth as time runs on. They split the error with the help of a known equation for the spread of the swarm, and they use the fast decay of its norms. In the end, the swarm's spread shrinks toward a single point at the global best value, in both swarm size and time.

Catalog record

CREATORErhan Bayraktar, Ibrahim Ekren, Hongyi Zhou
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE120 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-00582

Browse the full library · About the Tower of Babel