Table 5.
Classification metrics comparison for other preprocessing methodologies applied to the VRU system.
| Training | Test | |||||||
|---|---|---|---|---|---|---|---|---|
| Methodology | Accuracy | False Healthy | False Degraded |
F1 Score |
Accuracy | False Healthy | False Degraded |
F1 Score |
| AE (unsupervised) | 0.99 | 0.01 | 0.01 | 0.62 | 0.50 | 0.50 | 0.00 | 0.33 |
| AE-RF | 0.99 | 0.00 | 0.00 | 1.00 | 0.53 | 0.47 | 0.00 | 0.40 |
| AE-NN | 0.99 | 0.01 | 0.00 | 0.99 | 0.50 | 0.50 | 0.00 | 0.33 |
| PCA-RF | 1.00 | 0.00 | 0.00 | 1.00 | 0.57 | 0.43 | 0.00 | 0.57 |
| PCA-NN | 0.99 | 0.01 | 0.00 | 0.99 | 0.50 | 0.50 | 0.00 | 0.50 |
| AE-RF (Balanced) | 1.00 | 0.00 | 0.00 | 1.00 | 0.94 | 0.06 | 0.00 | 0.94 |
| AE-NN (Balanced) | 1.00 | 0.00 | 0.00 | 1.00 | 0.79 | 0.20 | 0.01 | 0.79 |
| PCA-RF (Balanced) | 1.00 | 0.00 | 0.00 | 1.00 | 0.98 | 0.02 | 0.00 | 0.98 |
| PCA-NN (Balanced) | 0.98 | 0.00 | 0.02 | 0.98 | 0.98 | 0.00 | 0.02 | 0.98 |