A new descent extension of DY conjugate gradient method based on DL approach

Abstract

Optimization problems are ways by which decisions are being made and these are found in all sectors of human endeavors. Conjugate gradient methods have been widely used as a result of their efficiency and low memory status in solving large-scale optimization problems. This article introduces an inexact line search scheme based algorithm. Analyses done showed that it satisfied decent property and converges globally. It also outperformed other methods considered when subjected to numerical test under Dolan and More performance profile tools.