Etude comparative théorique de deux fonctions noyau pour la programmation linéaire

Abstract

In this thesis, we are interested in the study of the improvement of the behavior of the algorithm by looking in particular at the algorithmic complexity via new techniques among which kernel functions. for this we gave theoretical comparative study of two kernel functions for linear programming ( Kernel function of logarithmic barrier terms, Kernel function with trigonometric barrier terms ), the theoretical comparative study prove that the kernel function of logarithmic barrier terms more efficient compared kernel function with trigonometric barrier terms.