Scholarlink

<p>The increasing use of concrete along with the emergence of innovative types of concrete necessitates in-depth knowledge regarding their performance. Given the inherent uncertainties, time constraints, and costs associated with traditional laboratory tests, the application of machine learning (ML) as a powerful technique of artificial intelligence (AI) has recently gained particular interest for predicting properties of concrete and optimizing concrete mixtures. Keeping that in mind, this review paper explores the application of ML models within the field of concrete technology, investigating various aspects. It includes the prediction of concrete properties, addressing classification challenges, and exploring advanced ML methodologies such as Automated ML, explainable AI, generative models, and counterfactual analysis. Furthermore, the paper emphasizes the critical importance of data preprocessing for optimizing the performance of these methods. In this regard, this paper serves as a comprehensive resource, providing researchers with a profound understanding of ML model development in the context of concrete technology.</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 Learning as an Innovative Engineering Tool for Controlling Concrete Performance: A Comprehensive Review

  • 期刊论文

  • 发表日期：

    2025年4月10日

查看全文
  • Fatemeh Mobasheri,
  • Masoud Hosseinpoor,
  • Ammar Yahia,
  • Farhad Pourkamali-Anaraki

摘要

The increasing use of concrete along with the emergence of innovative types of concrete necessitates in-depth knowledge regarding their performance. Given the inherent uncertainties, time constraints, and costs associated with traditional laboratory tests, the application of machine learning (ML) as a powerful technique of artificial intelligence (AI) has recently gained particular interest for predicting properties of concrete and optimizing concrete mixtures. Keeping that in mind, this review paper explores the application of ML models within the field of concrete technology, investigating various aspects. It includes the prediction of concrete properties, addressing classification challenges, and exploring advanced ML methodologies such as Automated ML, explainable AI, generative models, and counterfactual analysis. Furthermore, the paper emphasizes the critical importance of data preprocessing for optimizing the performance of these methods. In this regard, this paper serves as a comprehensive resource, providing researchers with a profound understanding of ML model development in the context of concrete technology.