An improved alternating inertial algorithm with adaptive step-sizes for equilibrium problems with data classification experiments

doi:10.1016/j.cam.2026.117393

An improved alternating inertial algorithm with adaptive step-sizes for equilibrium problems with data classification experiments

2026-02-01

0

PRE

AI

DOI：10.1016/j.cam.2026.117393

Original

Original request for help

Save

Abstract

En 中文

We propose an alternated inertial subgradient extragradient algorithm with non-monotonic adaptive step sizes and relaxation effects for finding solutions to equilibrium problems in real Hilbert spaces. Our algorithm uses a non-monotonic step size criterion, which allows the algorithm to adaptively adjust stepsizes during iterations without requiring the prior information of the Lipschitz constant of the bifunction. The Fej & eacute;r monotonicity of the even subsequence generated by the proposed algorithm with respect to the solution is recovered. Weak and linear convergence of the proposed algorithm is established under the condition that the involved bifunction is pseudomonotone and strongly pseudomonotone, respectively. Several numerical experiments, including data classification, are conducted to illustrate the computational performances of the proposed algorithm.

Keywords:

Equilibrium problem

Extragradient method

Alternated inertial method

Pseudomonotone operator

Data classification

An improved alternating inertial algorithm with adaptive step-sizes for equilibrium problems with data classification experiments

Abstract

Journal

Organization