Scholarlink

<p>In this study, a mobile application is developed for predicting and interpreting the compressive strength of high-performance concrete using machine learning and explainable artificial intelligence techniques. Four models—Linear Regression, Random Forest, Support Vector Regression, and XGBoost—are evaluated, with XGBoost achieving the best performance (RMSE = 3.25 MPa). Model transparency is ensured using SHAP based on the TreeSHAP algorithm. Global SHAP analysis identifies concrete age, cement content, and water content as the most influential features, with importance scores of 8.88, 6.69, and 4.70, respectively. Local SHAP explanations are further generated to interpret individual model predictions. To enable real-time local interpretability, the trained model and SHAP explainer are deployed on a server, while a lightweight mobile application serves as the user interface. The application transmits user inputs to the server and receives both predicted compressive strength values and the corresponding local SHAP explanations in real time. This architecture enables on-site usability, low device requirements, and real-time interpretability. The proposed system provides an accurate and transparent decision support tool for concrete strength assessment and demonstrates the practical potential of mobile-based explainable machine learning applications in civil engineering.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of a Mobile Application for Predicting and Interpreting High-Performance Concrete Compressive Strength

  • 期刊论文

  • 发表日期：

    2026年3月11日

查看全文
  • Mohsen Beigi

摘要

In this study, a mobile application is developed for predicting and interpreting the compressive strength of high-performance concrete using machine learning and explainable artificial intelligence techniques. Four models—Linear Regression, Random Forest, Support Vector Regression, and XGBoost—are evaluated, with XGBoost achieving the best performance (RMSE = 3.25 MPa). Model transparency is ensured using SHAP based on the TreeSHAP algorithm. Global SHAP analysis identifies concrete age, cement content, and water content as the most influential features, with importance scores of 8.88, 6.69, and 4.70, respectively. Local SHAP explanations are further generated to interpret individual model predictions. To enable real-time local interpretability, the trained model and SHAP explainer are deployed on a server, while a lightweight mobile application serves as the user interface. The application transmits user inputs to the server and receives both predicted compressive strength values and the corresponding local SHAP explanations in real time. This architecture enables on-site usability, low device requirements, and real-time interpretability. The proposed system provides an accurate and transparent decision support tool for concrete strength assessment and demonstrates the practical potential of mobile-based explainable machine learning applications in civil engineering.