Scholarlink

<p>Due to the vast compositional space, conventional “trial-and-error” and computational methods struggle to effectively uncover the composition-property relationships in high-entropy alloys (HEAs). In this study, we proposed an ensemble averaging-based machine learning (ML) strategy to predict the hardness of a diverse range of HEAs. Initially, we constructed a composition-hardness database of HEAs comprising over nine hundred data points, covering nearly thirty elements. Six key features affecting the hardness of HEAs were identified through a three-step feature selection process. A highly reliable ensemble model was subsequently developed to predict the hardness of HEAs. Furthermore, SHapley Additive exPlanation (SHAP) analysis was conducted to interpret the contributions of the selected features to the hardness of HEAs. This work presents an efficient approach for establishing high-predictability and generalizability ML models using an ensemble learning strategy, demonstrating promising potential in developing advanced HEAs.</p> Graphical Abstract <p></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

Ensemble Learning for Comprehensive Hardness Prediction of High-Entropy Alloys

  • 期刊论文

  • 发表日期：

    2025年11月8日

查看全文
  • Libin Zhang,
  • Taejoo Lee,
  • Yoon Suk Choi

摘要

Due to the vast compositional space, conventional “trial-and-error” and computational methods struggle to effectively uncover the composition-property relationships in high-entropy alloys (HEAs). In this study, we proposed an ensemble averaging-based machine learning (ML) strategy to predict the hardness of a diverse range of HEAs. Initially, we constructed a composition-hardness database of HEAs comprising over nine hundred data points, covering nearly thirty elements. Six key features affecting the hardness of HEAs were identified through a three-step feature selection process. A highly reliable ensemble model was subsequently developed to predict the hardness of HEAs. Furthermore, SHapley Additive exPlanation (SHAP) analysis was conducted to interpret the contributions of the selected features to the hardness of HEAs. This work presents an efficient approach for establishing high-predictability and generalizability ML models using an ensemble learning strategy, demonstrating promising potential in developing advanced HEAs.

Graphical Abstract