.png){kind=icon}
.png){kind=icon}
Return
Deep learning for high-entropy alloys: Phase prediction and feature interpretability through hyperparameter optimization
.png){kind=icon}
PRE
AI
DOI：10.1177/09544062251414915.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
High-entropy alloys (HEAs), characterized by multi-principal elements and high configurational entropy, exhibit exceptional mechanical, thermal, and corrosion-resistant properties, making them promising candidates for aerospace, energy, and biomedical applications. However, predicting their phase formation-single-phase solid solutions, multi-phase solid solutions, or intermetallic compounds-remains challenging due to the interplay of thermodynamic and atomic-scale parameters. This study leverages deep neural networks (DNNs) to predict HEA phases using six compositional features: entropy of mixing (Delta Smix), enthalpy of mixing (Delta Hmix), atomic size difference ( delta ), thermodynamic stability ( Omega ) parameter, valence electron concentration (VEC), and electronegativity difference (Delta chi). The optimized DNN architecture (three hidden layers with 128-128-64 neurons) achieved 78.6% accuracy and 78.5% F1-score, outperforming shallow models. SHapley Additive explanation (SHAP) analysis revealed Delta Hmix and Delta chi as dominant features, where exothermic mixing (negative Delta Hmix) and moderate Delta chi favored single-phase solid solution and Multiphase solid solution, while low Delta Smix and high delta promoted intermetallic phase formation. Thermodynamic stability ( Omega ) and VEC further distinguished phase regimes, with Omega >= 1.1 and VEC 8-10 favoring FCC/BCC solid solutions. Hyperparameter tuning highlighted the critical role of learning rate (optimal 0.001-0.0001) and model depth, where deeper networks (3 layers) enhanced performance but risked overfitting with limited data. These insights enable targeted alloy development, balancing entropy-driven stabilization and atomic-scale effects, and provide a framework for integrating computational tools in materials discovery.
Keywords:
high-entropy alloys (HEAs)
deep neural networks (DNNs)
hyperparameter optimization
SHapley Additive exPlanation (SHAP) analysis
phase prediction
Journal
IF:
1.7
Papers:
305
Citations:
1.0W