Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Practical machine learning techniques for estimating the splitting-tensile strength of recycled aggregate concrete
This paper deals with the development of two machine learning techniques, including K-Nearest Neighbors (KNN) and Random Forest (RF), for calculating the splitting-tensile strength of recycled aggregate concrete (RAC). Accordingly, a total of 134 experimental data points collected from the literature were employed to develop the KNN and RF models. The results demonstrated that the performance of the RF achieved high accuracy with correlation coefficients (R2) exceeding 0.94, and it can provide the best prediction when compared to the KNN model. Besides, the shapely additive explanations (SHAP) and sensitive analysis based on the RF model show that water, cement, and sand have the most significant impact on the splitting-tensile strength of RAC. Finally, a web application (WA) was developed to make the KNN and RF models more and more smart in predicting the splitting-tensile strength of RAC. Thanks to WA, civil engineering readily applied for predicting the splitting-tensile strength of RAC in the design process.
Practical machine learning techniques for estimating the splitting-tensile strength of recycled aggregate concrete
This paper deals with the development of two machine learning techniques, including K-Nearest Neighbors (KNN) and Random Forest (RF), for calculating the splitting-tensile strength of recycled aggregate concrete (RAC). Accordingly, a total of 134 experimental data points collected from the literature were employed to develop the KNN and RF models. The results demonstrated that the performance of the RF achieved high accuracy with correlation coefficients (R2) exceeding 0.94, and it can provide the best prediction when compared to the KNN model. Besides, the shapely additive explanations (SHAP) and sensitive analysis based on the RF model show that water, cement, and sand have the most significant impact on the splitting-tensile strength of RAC. Finally, a web application (WA) was developed to make the KNN and RF models more and more smart in predicting the splitting-tensile strength of RAC. Thanks to WA, civil engineering readily applied for predicting the splitting-tensile strength of RAC in the design process.
Practical machine learning techniques for estimating the splitting-tensile strength of recycled aggregate concrete
Asian J Civ Eng
Phan, Tan-Duy (Autor:in)
Asian Journal of Civil Engineering ; 24 ; 3689-3710
01.12.2023
22 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Tensile strength behaviour of recycled aggregate concrete
| Online Contents | 2015
Tensile strength behaviour of recycled aggregate concrete
| British Library Online Contents | 2015