Comparative Evaluation of Feed-Forward Neural Networks for Predicting Uniaxial Compressive Strength of Seybaplaya Carbonate Rock Cores

Accurate estimation of the uniaxial compressive strength (UCS) of carbonate rocks underpins safe design and stability assessment in karst-influenced geotechnical projects. This work presents a comprehensive evaluation of four feed-forward artificial neural network (ANN) architectures—radial basis fu...

Full description

Saved in:
Bibliographic Details
Main Authors: Jose W. Naal-Pech, Leonardo Palemón-Arcos, Youness El Hamzaoui
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/10/5609
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850257951536709632
author Jose W. Naal-Pech
Leonardo Palemón-Arcos
Youness El Hamzaoui
author_facet Jose W. Naal-Pech
Leonardo Palemón-Arcos
Youness El Hamzaoui
author_sort Jose W. Naal-Pech
collection DOAJ
description Accurate estimation of the uniaxial compressive strength (UCS) of carbonate rocks underpins safe design and stability assessment in karst-influenced geotechnical projects. This work presents a comprehensive evaluation of four feed-forward artificial neural network (ANN) architectures—radial basis function (RBF), Bayesian regularized (BR), scaled conjugate gradient (SCG), and Levenberg–Marquardt (LM)—to predict UCS from three readily measured variables: water content, interconnected porosity, and real density. Fifty core specimens from the Seybaplaya quarry in Campeche, Mexico, were split into training and testing subsets under uniform preprocessing. Each model’s predictive performance was assessed over 30 independent runs using mean absolute error, root mean squared error, and coefficient of determination, with statistical differences tested via nonparametric hypothesis testing. The RBF network achieved the highest median R<sup>2</sup> and significantly outperformed the other variants, while the BR model demonstrated robust generalization. SCG and LM converged faster and efficiently but with slightly lower accuracy. Sensitivity analysis identified interconnected porosity as the primary predictor of UCS. These results establish RBF-based ANNs with appropriate regularization and feature importance assessment as a novel, practical, and reliable framework for UCS prediction in heterogeneous carbonate formations.
format Article
id doaj-art-1876e169bc02424a895ea603672f5a8a
institution OA Journals
issn 2076-3417
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-1876e169bc02424a895ea603672f5a8a2025-08-20T01:56:17ZengMDPI AGApplied Sciences2076-34172025-05-011510560910.3390/app15105609Comparative Evaluation of Feed-Forward Neural Networks for Predicting Uniaxial Compressive Strength of Seybaplaya Carbonate Rock CoresJose W. Naal-Pech0Leonardo Palemón-Arcos1Youness El Hamzaoui2Engineering College, Carmen Autonomous University, Campus III, Avenida Central S/N, Esq. con Fracc. Mundo Maya, Carmen City C.P. 24115, Campeche, MexicoEngineering College, Carmen Autonomous University, Campus III, Avenida Central S/N, Esq. con Fracc. Mundo Maya, Carmen City C.P. 24115, Campeche, MexicoEngineering College, Carmen Autonomous University, Campus III, Avenida Central S/N, Esq. con Fracc. Mundo Maya, Carmen City C.P. 24115, Campeche, MexicoAccurate estimation of the uniaxial compressive strength (UCS) of carbonate rocks underpins safe design and stability assessment in karst-influenced geotechnical projects. This work presents a comprehensive evaluation of four feed-forward artificial neural network (ANN) architectures—radial basis function (RBF), Bayesian regularized (BR), scaled conjugate gradient (SCG), and Levenberg–Marquardt (LM)—to predict UCS from three readily measured variables: water content, interconnected porosity, and real density. Fifty core specimens from the Seybaplaya quarry in Campeche, Mexico, were split into training and testing subsets under uniform preprocessing. Each model’s predictive performance was assessed over 30 independent runs using mean absolute error, root mean squared error, and coefficient of determination, with statistical differences tested via nonparametric hypothesis testing. The RBF network achieved the highest median R<sup>2</sup> and significantly outperformed the other variants, while the BR model demonstrated robust generalization. SCG and LM converged faster and efficiently but with slightly lower accuracy. Sensitivity analysis identified interconnected porosity as the primary predictor of UCS. These results establish RBF-based ANNs with appropriate regularization and feature importance assessment as a novel, practical, and reliable framework for UCS prediction in heterogeneous carbonate formations.https://www.mdpi.com/2076-3417/15/10/5609uniaxial compressive strengthcarbonate rock coresartificial neural networksradial basis functionBayesian regularizationsensitivity analysis
spellingShingle Jose W. Naal-Pech
Leonardo Palemón-Arcos
Youness El Hamzaoui
Comparative Evaluation of Feed-Forward Neural Networks for Predicting Uniaxial Compressive Strength of Seybaplaya Carbonate Rock Cores
Applied Sciences
uniaxial compressive strength
carbonate rock cores
artificial neural networks
radial basis function
Bayesian regularization
sensitivity analysis
title Comparative Evaluation of Feed-Forward Neural Networks for Predicting Uniaxial Compressive Strength of Seybaplaya Carbonate Rock Cores
title_full Comparative Evaluation of Feed-Forward Neural Networks for Predicting Uniaxial Compressive Strength of Seybaplaya Carbonate Rock Cores
title_fullStr Comparative Evaluation of Feed-Forward Neural Networks for Predicting Uniaxial Compressive Strength of Seybaplaya Carbonate Rock Cores
title_full_unstemmed Comparative Evaluation of Feed-Forward Neural Networks for Predicting Uniaxial Compressive Strength of Seybaplaya Carbonate Rock Cores
title_short Comparative Evaluation of Feed-Forward Neural Networks for Predicting Uniaxial Compressive Strength of Seybaplaya Carbonate Rock Cores
title_sort comparative evaluation of feed forward neural networks for predicting uniaxial compressive strength of seybaplaya carbonate rock cores
topic uniaxial compressive strength
carbonate rock cores
artificial neural networks
radial basis function
Bayesian regularization
sensitivity analysis
url https://www.mdpi.com/2076-3417/15/10/5609
work_keys_str_mv AT josewnaalpech comparativeevaluationoffeedforwardneuralnetworksforpredictinguniaxialcompressivestrengthofseybaplayacarbonaterockcores
AT leonardopalemonarcos comparativeevaluationoffeedforwardneuralnetworksforpredictinguniaxialcompressivestrengthofseybaplayacarbonaterockcores
AT younesselhamzaoui comparativeevaluationoffeedforwardneuralnetworksforpredictinguniaxialcompressivestrengthofseybaplayacarbonaterockcores