Table 3 Evaluation metrics used to assess the performance and effectiveness of various regression models trained on two datasets for predicting Curie temperature
From: The northeast materials database for magnetic materials
Model | Dataset | R-squared (Test) | MAE (Test) | RMSE (Test) |
|---|---|---|---|---|
Random Forest | Original | 0.82 | 71K | 115K |
Balanced (Undersampled) | 0.85 | 65K | 104K | |
Ensembled Neural Network | Original | 0.84 | 60K | 110K |
Balanced (Undersampled) | 0.85 | 56K | 103K | |
XGBoost | Original | 0.84 | 62K | 107K |
Balanced (Undersampled) | 0.87 | 56K | 97K |
