A Monte Carlo Approach for Nonsmooth Convex Optimization via Proximal Splitting Algorithms

Nicholas Di, Eric C. Chi, Samy Wu Fung · 2025

Read A Monte Carlo Approach for Nonsmooth Convex Optimization via Proximal Splitting Algorithms by Nicholas Di, Eric C. Chi, Samy Wu Fung (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-2509-07914.txt.

Description

Nicholas Di and 2 co-authors widen the reach of a Monte Carlo trick for hard optimisation problems in a 2025 mathematics paper. Many first-order optimisation methods split a problem into pieces handled by proximal operators, mathematical steps that pull a point toward good values. Closed formulas for those steps exist only for limited function classes.

A recent method called HJ-Prox approximates the proximal step by random sampling, drawing on partial differential equation theory, with no formula needed. The authors prove it can be dropped into standard splitting schemes while keeping convergence guarantees. Nonsmooth problems, where the target has sharp corners, appear throughout imaging and machine learning, so a sampling-based step broadens where these trusted algorithms can be applied.

Catalog record

CREATORNicholas Di, Eric C. Chi, Samy Wu Fung
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-07914

Browse the full library · About the Tower of Babel