A new numerical scheme for solving linear optimization problems

Abstract

In this work, we analyze a logarithmic penalty method for linear optimization problems based on a novel approximation function. This function enables the computation of the step-size with ease and without significant complications. Numerical results are presented to highlight the efficiency and enhanced performance of the proposed method compared to classical line search techniques.