Table 8 Performance results of different classification techniques for the detection of PD on balanced and original datasets.
Techniques | Before SMOTE (original dataset) | After SMOTE (balanced dataset) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Acc | Sen. | Spec. | F1 Score | Training time (Sec.) | Acc. | Sen. | Spec. | F1 Score | Training time (Sec.) | |
RFT | 96.2 | 95.3 | 93.2 | 94.5 | 20 | 97.5 | 97 | 95 | 96.4 | 14.7 |
SVM | 95.2 | 93.4 | 92.2 | 93.4 | 25 | 96 | 94.5 | 94 | 95.5 | 20 |
K-NN | 94.2 | 91 | 92.7 | 90.5 | 22 | 95.7 | 92.4 | 93.5 | 92 | 16 |
NB | 92.3 | 90 | 91.3 | 87 | 20 | 93.2 | 91 | 92 | 89.5 | 14 |
LR | 90 | 87 | 82.5 | 86.3 | 25 | 91.1 | 89 | 84.5 | 88 | 19 |
