A NEW THREE-TERM CONJUGATE GRADIENT METHOD AND ITS APPLICATION IN TRAINING NEURAL NETWORK

Hosseini-meighan, Faeze Sadat; Ashrafi, Ali

doi:10.22055/jamm.2026.48572.2336

A NEW THREE-TERM CONJUGATE GRADIENT METHOD AND ITS APPLICATION IN TRAINING NEURAL NETWORK

Document Type : Original Paper

Authors

Faeze sadat Hosseini-meighan ¹

Ali Ashrafi ²

¹ Department of Mathematics, Faculty of Mathematical Sciences and Computer, Semnan University,

² Faculty of Mathematics, Statistics and Computer Science, Semnan University, Semnan, Iran

10.22055/jamm.2026.48572.2336

Abstract

Jiang et al. introduced a new family of three-term conjugate gradient tech niques that were theoretically effective and numerically ecient. Here, based on their study, we propose a new three-term conjugate gradient algorithm as a modified version of the Hastens-Stifel method. Numerical experiments are performed on a set of CUTER
test functions as well as on a neural network. They show that the proposed method has promising performance from a computational perspective.

Keywords

Nonlinear programming

unconstrained optimization

Three term conjugate gradient method

Neural network

Subjects