| About this Abstract |
| Meeting |
MS&T23: Materials Science & Technology
|
| Symposium
|
Materials Processing and Fundamental Understanding Based on Machine Learning and Data Informatics
|
| Presentation Title |
Machine Learning for Phase Prediction of High-entropy Alloys Assisted by Imbalance Learning |
| Author(s) |
Yoon Suk Choi, Libin Zhang, Dae-Geun Nam |
| On-Site Speaker (Planned) |
Libin Zhang |
| Abstract Scope |
Predicting the phase formation is a crucial step in novel high-entropy alloys (HEAs) design. Herein, we combined machine learning and imbalance learning algorithms to predict the phase structure in HEAs. In this work, we constructed an extensive database by collecting experimental data from the literature, and the key features affecting the phase formation of HEAs were filtered out by performing a three-step feature selection. Then, extreme gradient boosting (XGB) models were constructed to categorize the phases of HEAs with high accuracies. Moreover, we employed the Synthetic Minority Oversampling Technique (SMOTE) algorithm for data oversampling to address the data imbalance issue. It was found that imbalanced learning significantly improves the phase prediction, particularly for the minority class, without costing the overall prediction accuracy. |