Table 1 Used input features and short description for the machine learning models.
| Â | Used input features for each model | Short Description | |||
|---|---|---|---|---|---|
Chemical Composition | Process Condition | Thermodynamic Descriptor | Elemental Descriptor | ||
CD | ✔ | ✔ |  |  | Simple DNN model |
TD |  | ✔ | ✔ |  | Simple DNN model |
CTD | ✔ | ✔ | ✔ |  | Simple DNN model |
CD + CNN | ✔ | ✔ |  | ✔ | Concatenated DNN and CNN model |
TD + CNN | ✔ | ✔ | ✔ | ✔ | Concatenated DNN and CNN model |
CTD + CNN | ✔ | ✔ | ✔ | ✔ | Concatenated DNN and CNN model |
D + CNN | ✔ | ✔ |  | ✔ | Concatenated DNN and CNN model |
w/o T | ✔ | ✔ |  | ✔ | Ensemble model |
w/o CNN | ✔ | ✔ | ✔ |  | Ensemble model |
w/ T&C | ✔ | ✔ | ✔ | ✔ | Ensemble model |
XGBoost | ✔ | ✔ |  |  | XGBoost regressor model |
LR | ✔ | ✔ |  |  | Linear regressor model |
SVR | ✔ | ✔ |  |  | Support vector machine regressor |
DT | ✔ | ✔ |  |  | Decision tree regressor |
RF | ✔ | ✔ |  |  | Random forest regressor |
KNN | ✔ | ✔ |  |  | K-nearest neighbor regressor |
