Table 3.
Performance analysis of different classifiers on all extracted measurements.
| ML Classifier Name |
Best Parameters for Accuracy (If Exist) |
Accuracy before Applying K-Fold Cross Validation and PCA |
Accuracy after Applying K-Fold Cross Validation Only |
Accuracy after Applying PCA Only |
Accuracy after Applying K-Fold Cross Validation and PCA |
Recall Score |
Precis Ion |
|---|---|---|---|---|---|---|---|
| Logistic Regression (LR) |
- | 66.67% | 90.00% | (n_components = 20) 66.67% |
92.82% | 75.00% | 60.00% |
| K-Nearest Neighbours (K-NN) |
metric: ‘minkowski’, and n_neighbours = 5 |
77.78% | 55.00% | (n_components = 20) 77.78% |
97.11% | 75.00% | 75.00% |
| Support vector machine (SVM) |
C = 0.25, and kernel: ‘linear’ |
66.67% | 90.00% | (n_components = 20) 66.67% |
96.05% | 75.00% | 60.00% |
| Naive Bayes “GaussianNB” |
- | 77.78% | 70.00% | (n_components = 20) 77.78% |
81.14% | 75.00% | 75.00% |