Harnessing machine learning for predictive modelling of high entropy alloy phases
摘要
The application of classification-based machine-learning techniques offers a faster approach to designing high entropy alloys (HEAs). In this study, we have established a comprehensive framework for expedited phase identification within HEAs employing the ML approach by utilizing a distinct database. The trained model was subsequently employed to predict phases in HEA systems. Comparative analyses with experimental data corroborate the success of the ML models in accurately predicting phase transitions within HEAs across diverse composition profiles. This developed framework exhibits significant promise and applicability in the domain of material design, paving the way for the creation of novel materials.
Graphical abstract