Scholarlink

<p>In this study, machine intelligence (MI) models were developed to predict the compressive strength of concrete incorporating fly ash and blast furnace slag as supplementary cementitious materials. Key innovations include the integration of SHAP (SHapley Additive exPlanations) analysis for enhanced model interpretability, providing insights into the non-linear interactions among mix components and highlighting the most influential variables. In addition, a user-friendly graphical user interface (GUI) was developed to facilitate real-world applications, enabling efficient and accurate predictions for concrete mix optimization. A comprehensive database of 1030 concrete samples was used to train and validate advanced MI algorithms, including Adaptive Boosting (ADB), XGBoost (XGB), CatBoost (CATB), and Support Vector Regression (SVR), optimized with metaheuristic algorithms. Among these, CatBoost demonstrated superior performance, achieving an R<sup>2</sup> of 0.92 and a low RMSE of 0.03 during testing. The study aligns with the Sustainable Development Goals (SDGs) by demonstrating the dual benefits of improved concrete performance and reduced environmental impact through the use of supplementary materials. The findings provide valuable insights into the complex interactions between mixture components and their influence on compressive strength, offering a robust framework for designing high-performance eco-friendly concrete.</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

Machine intelligence models for predicting compressive strength of concrete incorporating fly ash and blast furnace slag

  • 期刊论文

  • 发表日期：

    2025年2月11日

查看全文
  • Abba Bashir,
  • Megha Gupta,
  • Sufyan Ghani

摘要

In this study, machine intelligence (MI) models were developed to predict the compressive strength of concrete incorporating fly ash and blast furnace slag as supplementary cementitious materials. Key innovations include the integration of SHAP (SHapley Additive exPlanations) analysis for enhanced model interpretability, providing insights into the non-linear interactions among mix components and highlighting the most influential variables. In addition, a user-friendly graphical user interface (GUI) was developed to facilitate real-world applications, enabling efficient and accurate predictions for concrete mix optimization. A comprehensive database of 1030 concrete samples was used to train and validate advanced MI algorithms, including Adaptive Boosting (ADB), XGBoost (XGB), CatBoost (CATB), and Support Vector Regression (SVR), optimized with metaheuristic algorithms. Among these, CatBoost demonstrated superior performance, achieving an R2 of 0.92 and a low RMSE of 0.03 during testing. The study aligns with the Sustainable Development Goals (SDGs) by demonstrating the dual benefits of improved concrete performance and reduced environmental impact through the use of supplementary materials. The findings provide valuable insights into the complex interactions between mixture components and their influence on compressive strength, offering a robust framework for designing high-performance eco-friendly concrete.