Table 2.
Classification accuracies for each preliminary feature set. The best classification result for each motion within each feature set is in bold.
| Feature Set | Motions | Classification Accuracy (%) | ||
|---|---|---|---|---|
| LDA | SVM | RF | ||
| FS1 (MAV, SSC, WL, ZC) |
EF | 62.3 | 60.5 | 70.3 |
| EE | 65.4 | 62.3 | 71.6 | |
| P | 69.1 | 67.9 | 67.3 | |
| S | 60.5 | 62.3 | 68.5 | |
| WF | 67.3 | 56.8 | 69.1 | |
| WE | 55.6 | 64.8 | 56.8 | |
| UD | 58.6 | 62.3 | 71.6 | |
| RD | 61.7 | 65.4 | 66.0 | |
| HC | 64.8 | 54.9 | 72.2 | |
| HO | 55.6 | 61.1 | 57.4 | |
| FS2 (RMS, AR2) |
EF | 59.9 | 57.4 | 67.9 |
| EE | 61.7 | 64.8 | 69.8 | |
| P | 67.3 | 49.4 | 71.6 | |
| S | 58.6 | 54.9 | 69.8 | |
| WF | 63.6 | 59.3 | 65.4 | |
| WE | 60.5 | 54.3 | 59.9 | |
| UD | 62.3 | 57.4 | 63.0 | |
| RD | 59.1 | 45.9 | 69.2 | |
| HC | 63.0 | 56.2 | 66.7 | |
| HO | 58.6 | 64.2 | 63.0 | |
| FS3 (MSA, MSF, MSS, MNPPS, MSD) |
EF | 74.1 | 61.1 | 64.8 |
| EE | 61.1 | 77.8 | 68.5 | |
| P | 72.2 | 63.0 | 61.1 | |
| S | 50.0 | 63.0 | 72.2 | |
| WF | 79.6 | 64.8 | 75.9 | |
| WE | 57.4 | 66.7 | 64.8 | |
| UD | 72.2 | 68.5 | 61.1 | |
| RD | 61.1 | 77.8 | 57.4 | |
| HC | 57.4 | 59.3 | 64.8 | |
| HO | 48.2 | 50.0 | 51.9 | |