Data augmentation and machine learning code in the prediction of Mg-based hydrogen storage alloy

This dataset includes augmented data related to the hydrogen storage capacity of Mg-based hydrogen storage alloys, as well as code used for data augmentation, model building, and model interpretation. The data augmentation code contains physical constraint augmentation, Gaussian noise augmentation, interpolation augmentation, and SMOTE oversampling augmentation. The model implementation includes code for SVR, MLP, RF, XGBoost, and the SHAP method.