Scholarlink

Concrete, being a primary construction material, occupies a prominent position in the current surge of construction activities. One crucial characteristic of concrete, particularly in the context of structural applications, is its flexural strength. In recent years, machine learning (ML) models have become increasingly valuable tools for predicting and enhancing concrete performance. Prior research has demonstrated the potential of ML models across a range of applications. The present study aims to predict the flexural strength in concrete made by incorporating recycled coarse and fine aggregates with a dataset consisting of 302 data points. In this study, ML models, such as K-nearest neighbours (KNN) and extra tree regression (ETR), have been utilized to generate predictive outcomes. In addition, this study also employs various data visualization techniques, including scatter plots, histograms, error analysis, coefficient of determination, and Taylor’s diagram. The results indicated that ETR outperforms KNN, as evidenced by a higher R-squared (R2) value of 0.84 for ETR compared to 0.41 for KNN. The aforementioned findings highlight the importance of ML models in accurately evaluating the flexural strength of concrete within the realm of sustainable construction practices.

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

Prediction of Flexural Strength of Concrete Manufactured by Integrating Construction and Demolition Waste: A Comparative Analysis of Machine Learning Approaches

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Rajwinder Singh,
  • Mahesh Patel

摘要

Concrete, being a primary construction material, occupies a prominent position in the current surge of construction activities. One crucial characteristic of concrete, particularly in the context of structural applications, is its flexural strength. In recent years, machine learning (ML) models have become increasingly valuable tools for predicting and enhancing concrete performance. Prior research has demonstrated the potential of ML models across a range of applications. The present study aims to predict the flexural strength in concrete made by incorporating recycled coarse and fine aggregates with a dataset consisting of 302 data points. In this study, ML models, such as K-nearest neighbours (KNN) and extra tree regression (ETR), have been utilized to generate predictive outcomes. In addition, this study also employs various data visualization techniques, including scatter plots, histograms, error analysis, coefficient of determination, and Taylor’s diagram. The results indicated that ETR outperforms KNN, as evidenced by a higher R-squared (R2) value of 0.84 for ETR compared to 0.41 for KNN. The aforementioned findings highlight the importance of ML models in accurately evaluating the flexural strength of concrete within the realm of sustainable construction practices.