Explainable Prediction of UHPC Tensile Strength Using Machine Learning with Engineered Features and Multi-Algorithm Comparative Evaluation
Open Access
- 6 September 2025
- Vol. 15 (17) , 3217
Abstract
To explore a direct predictive model for the tensile strength of ultra-high-performance concrete (UHPC), machine learning (ML) algorithms are presented. Initially, a database comprising 178 samples of UHPC tensile strength with varying parameters is established. Then, feature engineering strategies are proposed to optimize the robustness of ML models under a small-sample condition. Further, the performance and efficiency of algorithms are compared under default hyperparameters and hyperparameter tuning, respectively. Moreover, the utilization of SHapley Additive exPlanations (SHAP) enables the analysis of the relationships between UHPC tensile strength and its influencing factors. The quantitative analysis results indicate that ensemble algorithms exhibit superior performance, indicated by R2 values of above 0.92, under default hyperparameters. After hyperparameter tuning, both conventional and ensemble models achieve R2 values exceeding 0.94. However, Bayesian ridge regression (BRR) consistently demonstrates a suboptimal performance, irrespective of hyperparameter tuning. Notably, Categorical Boosting (CatBoost) requires a substantial duration of 1208 s, which is notably more time-consuming than that of other algorithms. The most influential feature identified is fiber reinforcement index with a contribution of 37.5%, followed by the water-to-cement ratio, strain rate, and cross-sectional size. The nonlinear relationship between UHPC tensile strength and the top four factors is visualized, and the critical thresholds are identified.Keywords
Funding Information
- Natural Science Foundation of Hunan Province, China (2025JJ70224, 2023JJ30216, 23B0576, 2023GZ2007)
- Research Foundation of Education Department of Hunan Province, China (2025JJ70224, 2023JJ30216, 23B0576, 2023GZ2007)
- Science and Technology Plan Project of Shaoyang City (2025JJ70224, 2023JJ30216, 23B0576, 2023GZ2007)
This publication has 74 references indexed in Scilit:
- Optimized decision tree algorithms to estimate ultimate strain of concrete wrapped by aramid fiber-reinforced polymerMultiscale and Multidisciplinary Modeling, Experiments and Design, 2025
- Machine and deep learning-based prediction of flexural moment capacity of ultra-high performance concrete beams with/out steel fiberAsian Journal of Civil Engineering, 2024
- Utilizing ensemble learning in the classifications of ductile and brittle failure modes of UHPC strengthened RC membersArchives of Civil and Mechanical Engineering, 2024
- Splitting tensile strength prediction of sustainable high-performance concrete using machine learning techniquesEnvironmental Science and Pollution Research, 2022
- Simplified analytical model to assess key factors influenced by fiber alignment and their effect on tensile performance of UHPCCement and Concrete Composites, 2022
- Assessment of compressive strength of Ultra-high Performance Concrete using deep machine learning techniquesApplied Soft Computing, 2020
- Damage investigation of ultra high performance concrete under direct tensile test using acoustic emission techniquesCement and Concrete Composites, 2018
- Fiber content and curing time effect on the tensile characteristics of ultra high performance fiber reinforced concreteStructural Concrete, 2017
- Strain rate dependent properties of ultra high performance fiber reinforced concrete (UHP-FRC) under tensionCement and Concrete Composites, 2015
- Properties of strain hardening ultra high performance fiber reinforced concrete (UHP-FRC) under direct tensile loadingCement and Concrete Composites, 2014