Research Article
Position-Based Feature Selection for Body Sensors regarding Daily Living Activity Recognition
Table 4
Number of selected features and accuracy rates depending on the sensor positions and feature selection methods.
| Number | Sensor position | Feature selection method | Number of selected features | Accuracy |
| 1 | Ankle | SFFS | 56 | 91.46% | Relief-F | 32 | 93.16% | Laplacian Score | 29 | 96.85% |
| 2 | Chest | SFFS | 87 | 95.62% | Relief-F | 77 | 95.11% | Laplacian Score | 36 | 95.62% |
| 3 | Hip | SFFS | 32 | 96.08% | Relief-F | 52 | 95.01% | Laplacian Score | 29 | 96.91% |
| 4 | Wrist | SFFS | 124 | 91.60% | Relief-F | 73 | 92.75% | Laplacian Score | 44 | 95.24% |
|
|