A New Method to Characterize Limited Material Datasets of High-Entropy Alloys Based on the Feature Engineering and Machine Learning

Abstract

This paper proposed a new method for characterizing limited material data of high-entropy alloys (HEAs) based on the feature engineering and machine learning (ML). The descriptor dimensionality is augmented from original small dimension to a high dimension by non-linear combination based on feature engineering to characterize this kind material. To avoid overfitting, we carried out 5-fold cross-validation to evaluate the generalization performance of the model. The results showed that this method could achieve higher accuracy in predicting the phase formation of HEAs. Except the prediction of HEAs, this method also can be applied to other materials with limited data.

Keywords: High-entropy alloy, Phase transformations, Machine learning, Feature engineering