Data-Efficient Training of Machine Learning Interatomic Potential for MAX-Phase Synthesizability Prediction

Key Points

• To develop a data-efficient machine learning model for predicting the synthesizability of high-entropy MAX phases.
• Utilized machine learning techniques to refine interatomic potentials.
• Employed data-efficient training methodologies to improve model accuracy.
• Analyzed the relationship between MAX phase features and synthetic outcomes.
• Demonstrated a high accuracy in predicting the synthesizability of HE-MAX phases.
• Showed improved data efficiency in training compared to conventional methods.
• Identified key features that influence the synthesis of MAX phases.