Table 3 The AUC values of RF and SVM models constructed using different feature sets at the first stage
Feature sets | AUC | |
|---|---|---|
RF (class weight balanced) | SVM (kernel function) | |
AAC | 0.68 | 0.63 (Polynomial kernel) |
AAindex | 0.69 | 0.69 (Radial basis function kernel with grid search hyperparameter tuning) |
ACC | 0.71 | 0.64 (Radial basis function kernel) |
BINARY | 0.59 | 0.71 (Polynomial kernel) |
BLOSUM62 | 0.68 | 0.74 (Radial basis function kernel) |
CKSAAP | 0.66 | 0.63 (Polynomial kernel) |
DISOPRED | 0.54 | 0.55 (Linear kernel) |
PSIPRED | 0.62 | 0.60 (Polynomial kernel) |
PSSM | 0.73 | 0.71 (Polynomial kernel) |
Selected features (mRMR+forward consequential elimination) | 0.75 | 0.72 (Linear kernel) |