A THREE-TERM CONJUGATE GRADIENT METHOD FOR LARGE-SCALE MINIMIZATION IN ARTIFICIAL NEURAL NETWORKS

Conjugate Gradient (CG) methods are widely used for solving unconstrained optimization problems due to their efficiency and low memory requirements. However, standard CG methods may not always guarantee sufficient descent condition, which can impact their robustness and convergence behavior. Additio...

Full description

Saved in:
Bibliographic Details
Main Authors: Umar A Omesa, Muhammad Y. Waziri, Issam A. R. Moghrabi, Sulaiman M. Ibrahim, Gudu E B, Fakai S L, Rabiu Bashir Yunus, Elissa Nadia Madi
Format: Article
Language:English
Published: Universitas Pattimura 2025-07-01
Series:Barekeng
Subjects:
Online Access:https://ojs3.unpatti.ac.id/index.php/barekeng/article/view/17704
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849770771399835648
author Umar A Omesa
Muhammad Y. Waziri
Issam A. R. Moghrabi
Sulaiman M. Ibrahim
Gudu E B
Fakai S L
Rabiu Bashir Yunus
Elissa Nadia Madi
author_facet Umar A Omesa
Muhammad Y. Waziri
Issam A. R. Moghrabi
Sulaiman M. Ibrahim
Gudu E B
Fakai S L
Rabiu Bashir Yunus
Elissa Nadia Madi
author_sort Umar A Omesa
collection DOAJ
description Conjugate Gradient (CG) methods are widely used for solving unconstrained optimization problems due to their efficiency and low memory requirements. However, standard CG methods may not always guarantee sufficient descent condition, which can impact their robustness and convergence behavior. Additionally, their effectiveness in training artificial neural networks (ANNs) remains an area of interest. In response, this paper presents a three-term conjugate gradient (CG) method for unconstrained optimization problems. The new parameter is formulated so that the search direction satisfies the sufficient descent condition. The global convergence result of the new algorithm is discussed under suitable assumptions. To evaluate the performance of the new method we considered some standard test problems for unconstrained optimization and applied the proposed method to train different ANNs on some benchmark data sets contained in the NN toolbox. The experimental results show that performance is encouraging for both unconstrained minimization test problems and in training neural networks.
format Article
id doaj-art-cf0eae78f5814e71a99014ef942b1cbc
institution DOAJ
issn 1978-7227
2615-3017
language English
publishDate 2025-07-01
publisher Universitas Pattimura
record_format Article
series Barekeng
spelling doaj-art-cf0eae78f5814e71a99014ef942b1cbc2025-08-20T03:02:54ZengUniversitas PattimuraBarekeng1978-72272615-30172025-07-011931973198810.30598/barekengvol19iss3pp1973-198817704A THREE-TERM CONJUGATE GRADIENT METHOD FOR LARGE-SCALE MINIMIZATION IN ARTIFICIAL NEURAL NETWORKSUmar A Omesa0Muhammad Y. Waziri1Issam A. R. Moghrabi2Sulaiman M. Ibrahim3Gudu E B4Fakai S L5Rabiu Bashir Yunus6Elissa Nadia Madi7Department of Mathematics, College of Science, Federal University of Agriculture, NigeriaDepartment of Mathematics, Faculty of Physical Sciences, Bayero University, NigeriaDepartment of Information Systems and Technology, Kuwait Technical College, KuwaitFaculty of Arts and Education, Sohar University, OmanDepartment of Mathematics, College of Science, Federal University of Agriculture, NigeriaDepartment of Mathematics, College of Science, Federal University of Agriculture, NigeriaDepartment of Mathematics, Aliko Dangote University of Science and Technology, NigeriaFaculty of Informatics and Computing, Universiti Sultan Zainal Abidin, Terengganu, MalaysiaConjugate Gradient (CG) methods are widely used for solving unconstrained optimization problems due to their efficiency and low memory requirements. However, standard CG methods may not always guarantee sufficient descent condition, which can impact their robustness and convergence behavior. Additionally, their effectiveness in training artificial neural networks (ANNs) remains an area of interest. In response, this paper presents a three-term conjugate gradient (CG) method for unconstrained optimization problems. The new parameter is formulated so that the search direction satisfies the sufficient descent condition. The global convergence result of the new algorithm is discussed under suitable assumptions. To evaluate the performance of the new method we considered some standard test problems for unconstrained optimization and applied the proposed method to train different ANNs on some benchmark data sets contained in the NN toolbox. The experimental results show that performance is encouraging for both unconstrained minimization test problems and in training neural networks.https://ojs3.unpatti.ac.id/index.php/barekeng/article/view/17704conjugate gradientglobal convergenceneural networkssufficient descent conditionunconstrained optimization
spellingShingle Umar A Omesa
Muhammad Y. Waziri
Issam A. R. Moghrabi
Sulaiman M. Ibrahim
Gudu E B
Fakai S L
Rabiu Bashir Yunus
Elissa Nadia Madi
A THREE-TERM CONJUGATE GRADIENT METHOD FOR LARGE-SCALE MINIMIZATION IN ARTIFICIAL NEURAL NETWORKS
Barekeng
conjugate gradient
global convergence
neural networks
sufficient descent condition
unconstrained optimization
title A THREE-TERM CONJUGATE GRADIENT METHOD FOR LARGE-SCALE MINIMIZATION IN ARTIFICIAL NEURAL NETWORKS
title_full A THREE-TERM CONJUGATE GRADIENT METHOD FOR LARGE-SCALE MINIMIZATION IN ARTIFICIAL NEURAL NETWORKS
title_fullStr A THREE-TERM CONJUGATE GRADIENT METHOD FOR LARGE-SCALE MINIMIZATION IN ARTIFICIAL NEURAL NETWORKS
title_full_unstemmed A THREE-TERM CONJUGATE GRADIENT METHOD FOR LARGE-SCALE MINIMIZATION IN ARTIFICIAL NEURAL NETWORKS
title_short A THREE-TERM CONJUGATE GRADIENT METHOD FOR LARGE-SCALE MINIMIZATION IN ARTIFICIAL NEURAL NETWORKS
title_sort three term conjugate gradient method for large scale minimization in artificial neural networks
topic conjugate gradient
global convergence
neural networks
sufficient descent condition
unconstrained optimization
url https://ojs3.unpatti.ac.id/index.php/barekeng/article/view/17704
work_keys_str_mv AT umaraomesa athreetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT muhammadywaziri athreetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT issamarmoghrabi athreetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT sulaimanmibrahim athreetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT gudueb athreetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT fakaisl athreetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT rabiubashiryunus athreetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT elissanadiamadi athreetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT umaraomesa threetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT muhammadywaziri threetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT issamarmoghrabi threetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT sulaimanmibrahim threetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT gudueb threetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT fakaisl threetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT rabiubashiryunus threetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks
AT elissanadiamadi threetermconjugategradientmethodforlargescaleminimizationinartificialneuralnetworks