Table 3 Hyperparameter tuning of the SVM classifier using RBF kernel for optimal PQD classification performance.
From: Power system security and protection considering the integration of new energy power plants
Parameter | Description | Tested range | Optimal value | Tuning method |
|---|---|---|---|---|
C | Penalty parameter for misclassification | [0.1, 1, 10, 100] | 10 | Grid search + cross-validation |
γ | Kernel coefficient for RBF kernel | [0.01, 0.1, 1, 10] | 0.1 | Grid search + cross-validation |
Kernel | Type of kernel used in SVM | {Linear, polynomial, RBF} | RBF | Empirical evaluation |
CV folds | Number of folds for cross-validation | 5 | 5 | Fixed |