Table 4.
Overall accuracy (Acc), specificity (true negative rate (TNR)), sensitivity/recall (true positive rate (TPR)) and precision (positive predictive value (PPV)) values of the posture recognition when using the four different kernel functions considering the trunk–lower limb (T-Ll) and upper body (UpB) datasets in the RL task.
| RL Task | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 kg | 2 kg | 5 kg | |||||||||||
| SVM kernels | Acc | TNR | TPR | PPV | Acc | TNR | TPR | PPV | Acc | TNR | TPR | PPV | |
| T-Ll | Linear | 99.4 | 98.7 | 100 | 98.7 | 99.4 | 98.7 | 100 | 98.7 | 98.7 | 100 | 97.4 | 100 |
| Quadratic | 98.1 | 98.7 | 97.4 | 98.7 | 99.4 | 98.7 | 100 | 98.7 | 94.9 | 93.6 | 96.2 | 93.8 | |
| Cubic | 97.4 | 98.7 | 96.2 | 98.7 | 99.4 | 98.7 | 100 | 98.7 | 95.5 | 93.6 | 97.4 | 93.8 | |
| Gaussian | 98.1 | 97.4 | 98.7 | 97.5 | 98.7 | 97.4 | 100 | 97.5 | 98.1 | 98.7 | 97.4 | 98.7 | |
| UpB | Linear | 72.4 | 67.9 | 76.9 | 70.6 | 67.9 | 69.2 | 66.7 | 68.4 | 70.5 | 82.1 | 59.0 | 76.7 |
| Quadratic | 71.2 | 62.8 | 79.5 | 68.1 | 74.4 | 70.5 | 78.2 | 72.6 | 76.3 | 73.1 | 79.5 | 74.7 | |
| Cubic | 78.8 | 83.3 | 74.4 | 81.7 | 67.9 | 69.2 | 66.7 | 68.4 | 68.6 | 59.0 | 78.2 | 65.6 | |
| Gaussian | 83.3 | 78.2 | 88.5 | 80.2 | 76.9 | 80.8 | 73.1 | 79.2 | 71.2 | 65.4 | 76.9 | 69.0 | |