EUROPT 2025
Abstract Submission

EURO-Online login

40. High-Order Forward-Backward Methods for Nonconvex Optimization

CANCELED

Invited abstract in session MD-1: Smoothing techniques for nonsmooth optimization, stream Nonsmooth and nonconvex optimization.

Monday, 16:30-18:30

Authors (first author is the speaker)

1. Alireza Kabgani
Department of Mathematics, University of Antwerp
2. Masoud Ahookhosh
Department of Mathematics, University of Antwerp

Abstract

This talk presents high-order forward-backward splitting mappings and their associated envelopes, exploring their properties to design efficient numerical algorithms for composite nonsmooth nonconvex optimization problems. We establish global and linear convergence guarantees under the Kurdyka-Ɓojasiewicz inequality, supported by rigorous theoretical analysis. The effectiveness of these algorithms is demonstrated through numerical experiments, highlighting their potential in tackling challenging nonconvex optimization tasks.

Keywords

Status: accepted


Back to the list of papers