New SHapley Additive ExPlanations (SHAP) Approach to Evaluate the Raw Materials Interactions of Steel-Fiber-Reinforced Concrete
Recently, artificial intelligence (AI) approaches have gained the attention of researchers in the civil engineering field for estimating the mechanical characteristics of concrete to save the effort, time, and cost of researchers. Consequently, the current research focuses on assessing steel-fiber-r...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-09-01
|
Series: | Materials |
Subjects: | |
Online Access: | https://www.mdpi.com/1996-1944/15/18/6261 |
_version_ | 1797485415155892224 |
---|---|
author | Madiha Anjum Kaffayatullah Khan Waqas Ahmad Ayaz Ahmad Muhammad Nasir Amin Afnan Nafees |
author_facet | Madiha Anjum Kaffayatullah Khan Waqas Ahmad Ayaz Ahmad Muhammad Nasir Amin Afnan Nafees |
author_sort | Madiha Anjum |
collection | DOAJ |
description | Recently, artificial intelligence (AI) approaches have gained the attention of researchers in the civil engineering field for estimating the mechanical characteristics of concrete to save the effort, time, and cost of researchers. Consequently, the current research focuses on assessing steel-fiber-reinforced concrete (SFRC) in terms of flexural strength (FS) prediction by employing delicate AI techniques as well as to predict the raw material interaction that is still a research gap. In this study, the FS of SFRC is estimated by deploying supervised machine learning (ML) techniques, such as DT-Gradient Boosting, DT-XG Boost, DT-AdaBoost, and DT-Bagging. In addition to that, the performance model is also evaluated by using R<sup>2</sup>, root mean square error (RMSE), and mean absolute error (MAE). Furthermore, the k-fold cross-validation method is also applied to validate the model’s performance. It is observed that DT-Bagging with an R<sup>2</sup> value of 0.95 is superior to DT-XG Boost, DT-Gradient Boosting, and DT-AdaBoost. Lesser error MAE and RMSE and higher R<sup>2</sup> values for the DT-Bagging model show the enhanced performance of the model compared to the other ensembled approaches. Considerable conservation of time, effort, and cost can be made by applying ML techniques to predict concrete properties. The evaluation of the outcome depicts that the estimated results of DT-Bagging are closer to the experimental results, indicating the accurate estimation of SFRC flexural strength. It is further revealed from the SHapley Additive exPlanations (SHAP) study that the volumetric content of steel fiber highly and positively influences the FS of SFRC. |
first_indexed | 2024-03-09T23:18:28Z |
format | Article |
id | doaj.art-57aef61d36874f76aa87b349e8f2c842 |
institution | Directory Open Access Journal |
issn | 1996-1944 |
language | English |
last_indexed | 2024-03-09T23:18:28Z |
publishDate | 2022-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Materials |
spelling | doaj.art-57aef61d36874f76aa87b349e8f2c8422023-11-23T17:30:32ZengMDPI AGMaterials1996-19442022-09-011518626110.3390/ma15186261New SHapley Additive ExPlanations (SHAP) Approach to Evaluate the Raw Materials Interactions of Steel-Fiber-Reinforced ConcreteMadiha Anjum0Kaffayatullah Khan1Waqas Ahmad2Ayaz Ahmad3Muhammad Nasir Amin4Afnan Nafees5Department of Computer Engineering, College of Computer Science and Information, Technology, King Faisal University, Al-Ahsa 31982, Saudi ArabiaDepartment of Civil and Environmental Engineering, College of Engineering, King Faisal University, Al-Ahsa 31982, Saudi ArabiaDepartment of Civil Engineering, COMSATS University Islamabad, Abbottabad 22060, PakistanMaREI Centre, Ryan Institute and School of Engineering, College of Science and Engineering, National University of Ireland Galway, H91 TK33 Galway, IrelandDepartment of Civil and Environmental Engineering, College of Engineering, King Faisal University, Al-Ahsa 31982, Saudi ArabiaDepartment of Civil Engineering, COMSATS University Islamabad, Abbottabad 22060, PakistanRecently, artificial intelligence (AI) approaches have gained the attention of researchers in the civil engineering field for estimating the mechanical characteristics of concrete to save the effort, time, and cost of researchers. Consequently, the current research focuses on assessing steel-fiber-reinforced concrete (SFRC) in terms of flexural strength (FS) prediction by employing delicate AI techniques as well as to predict the raw material interaction that is still a research gap. In this study, the FS of SFRC is estimated by deploying supervised machine learning (ML) techniques, such as DT-Gradient Boosting, DT-XG Boost, DT-AdaBoost, and DT-Bagging. In addition to that, the performance model is also evaluated by using R<sup>2</sup>, root mean square error (RMSE), and mean absolute error (MAE). Furthermore, the k-fold cross-validation method is also applied to validate the model’s performance. It is observed that DT-Bagging with an R<sup>2</sup> value of 0.95 is superior to DT-XG Boost, DT-Gradient Boosting, and DT-AdaBoost. Lesser error MAE and RMSE and higher R<sup>2</sup> values for the DT-Bagging model show the enhanced performance of the model compared to the other ensembled approaches. Considerable conservation of time, effort, and cost can be made by applying ML techniques to predict concrete properties. The evaluation of the outcome depicts that the estimated results of DT-Bagging are closer to the experimental results, indicating the accurate estimation of SFRC flexural strength. It is further revealed from the SHapley Additive exPlanations (SHAP) study that the volumetric content of steel fiber highly and positively influences the FS of SFRC.https://www.mdpi.com/1996-1944/15/18/6261steel fiberbuilding materialflexural strengthfibersconcretemortar |
spellingShingle | Madiha Anjum Kaffayatullah Khan Waqas Ahmad Ayaz Ahmad Muhammad Nasir Amin Afnan Nafees New SHapley Additive ExPlanations (SHAP) Approach to Evaluate the Raw Materials Interactions of Steel-Fiber-Reinforced Concrete Materials steel fiber building material flexural strength fibers concrete mortar |
title | New SHapley Additive ExPlanations (SHAP) Approach to Evaluate the Raw Materials Interactions of Steel-Fiber-Reinforced Concrete |
title_full | New SHapley Additive ExPlanations (SHAP) Approach to Evaluate the Raw Materials Interactions of Steel-Fiber-Reinforced Concrete |
title_fullStr | New SHapley Additive ExPlanations (SHAP) Approach to Evaluate the Raw Materials Interactions of Steel-Fiber-Reinforced Concrete |
title_full_unstemmed | New SHapley Additive ExPlanations (SHAP) Approach to Evaluate the Raw Materials Interactions of Steel-Fiber-Reinforced Concrete |
title_short | New SHapley Additive ExPlanations (SHAP) Approach to Evaluate the Raw Materials Interactions of Steel-Fiber-Reinforced Concrete |
title_sort | new shapley additive explanations shap approach to evaluate the raw materials interactions of steel fiber reinforced concrete |
topic | steel fiber building material flexural strength fibers concrete mortar |
url | https://www.mdpi.com/1996-1944/15/18/6261 |
work_keys_str_mv | AT madihaanjum newshapleyadditiveexplanationsshapapproachtoevaluatetherawmaterialsinteractionsofsteelfiberreinforcedconcrete AT kaffayatullahkhan newshapleyadditiveexplanationsshapapproachtoevaluatetherawmaterialsinteractionsofsteelfiberreinforcedconcrete AT waqasahmad newshapleyadditiveexplanationsshapapproachtoevaluatetherawmaterialsinteractionsofsteelfiberreinforcedconcrete AT ayazahmad newshapleyadditiveexplanationsshapapproachtoevaluatetherawmaterialsinteractionsofsteelfiberreinforcedconcrete AT muhammadnasiramin newshapleyadditiveexplanationsshapapproachtoevaluatetherawmaterialsinteractionsofsteelfiberreinforcedconcrete AT afnannafees newshapleyadditiveexplanationsshapapproachtoevaluatetherawmaterialsinteractionsofsteelfiberreinforcedconcrete |