Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations

This paper introduces a novel multilayer neural network technique to solve partial differential equations with non-integer derivatives (FPDEs). The proposed model is a deep feed-forward multiple layer neural network (DFMLNN) that is trained using advanced optimization approaches, namely adaptive mom...

Full description

Bibliographic Details
Main Authors: Ali, Amina Hassan, Senu, Norazak, Ahmadian, Ali
Format: Article
Published: John Wiley and Sons 2025
Online Access:http://psasir.upm.edu.my/id/eprint/120691/
_version_ 1848868218167361536
author Ali, Amina Hassan
Senu, Norazak
Ahmadian, Ali
author_facet Ali, Amina Hassan
Senu, Norazak
Ahmadian, Ali
author_sort Ali, Amina Hassan
building UPM Institutional Repository
collection Online Access
description This paper introduces a novel multilayer neural network technique to solve partial differential equations with non-integer derivatives (FPDEs). The proposed model is a deep feed-forward multiple layer neural network (DFMLNN) that is trained using advanced optimization approaches, namely adaptive moment estimation (Adam) and limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS), which integrate neural networks. First, the Adam method is employed for training, and then the model is further improved using L-BFGS. The Laplace transform is used, concentrating on the Caputo fractional derivative, to approximate the FPDE. The efficacy of this strategy is confirmed through rigorous testing, which involves making predictions and comparing the outcomes with exact solutions. The results illustrate that this combined approach greatly improves both precision and effectiveness. This proposed multilayer neural network offers a robust and reliable framework for solving FPDEs.
first_indexed 2025-11-15T14:48:54Z
format Article
id upm-120691
institution Universiti Putra Malaysia
institution_category Local University
last_indexed 2025-11-15T14:48:54Z
publishDate 2025
publisher John Wiley and Sons
recordtype eprints
repository_type Digital Repository
spelling upm-1206912025-10-08T06:19:18Z http://psasir.upm.edu.my/id/eprint/120691/ Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations Ali, Amina Hassan Senu, Norazak Ahmadian, Ali This paper introduces a novel multilayer neural network technique to solve partial differential equations with non-integer derivatives (FPDEs). The proposed model is a deep feed-forward multiple layer neural network (DFMLNN) that is trained using advanced optimization approaches, namely adaptive moment estimation (Adam) and limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS), which integrate neural networks. First, the Adam method is employed for training, and then the model is further improved using L-BFGS. The Laplace transform is used, concentrating on the Caputo fractional derivative, to approximate the FPDE. The efficacy of this strategy is confirmed through rigorous testing, which involves making predictions and comparing the outcomes with exact solutions. The results illustrate that this combined approach greatly improves both precision and effectiveness. This proposed multilayer neural network offers a robust and reliable framework for solving FPDEs. John Wiley and Sons 2025 Article PeerReviewed Ali, Amina Hassan and Senu, Norazak and Ahmadian, Ali (2025) Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations. International Journal of Numerical Modelling: Electronic Networks, Devices and Fields, 38 (4). art. no. e70073. ISSN 0894-3370; eISSN: 1099-1204 https://onlinelibrary.wiley.com/doi/10.1002/jnm.70073 10.1002/jnm.70073
spellingShingle Ali, Amina Hassan
Senu, Norazak
Ahmadian, Ali
Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations
title Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations
title_full Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations
title_fullStr Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations
title_full_unstemmed Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations
title_short Multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations
title_sort multilayer neural networks enhanced with hybrid methods for solving fractional partial differential equations
url http://psasir.upm.edu.my/id/eprint/120691/
http://psasir.upm.edu.my/id/eprint/120691/
http://psasir.upm.edu.my/id/eprint/120691/