A Modified Spectral Methods for Large-Scale UnconStrained

Abstract

A modified spectral methods for solving unconstrained optimization problems based on the formulae are derived which are given in [4, 5]. The proposed methods satisfied the descent condition. Moreover, we prove that the new spectral methods are globally convergent. The Numerical results show that the proposed methods effective by comparing with the FR-method.

Keywords

• Conjugate gradient, Spectral conjugate gradient, Global convergence, Numerical results.