Research Article
Mobile Edge Computing Enabled Efficient Communication Based on Federated Learning in Internet of Medical Things
Table 3
Classification accuracy of every subject in arm swing and postural normal tremor.
| Subject | KNN | SVM | RF | DL | FT-IoMT Health | Upper bound |
| Arm swing | | 37.2 | 41.5 | 45.1 | 50.2 | 74.7 | 87.9 | | 45.3 | 47.6 | 49.2 | 58.5 | 91.2 | 99.4 | | 56.2 | 55.7 | 53.4 | 63.6 | 86.4 | 87.5 | | 63.8 | 62.0 | 56.7 | 69.4 | 94.6 | 100 | | 84.9 | 73.6 | 66.6 | 71.3 | 85.7 | 88.4 | AVG | 57.5 | 56.0 | 54.2 | 62.6 | 86.5 | 92.6 | Postural tremor | | 51.3 | 46.5 | 58.3 | 46.2 | 84.3 | 86.2 | | 52.5 | 58.7 | 56.9 | 60.4 | 75.8 | 85.9 | | 64.8 | 54.1 | 56.1 | 58.7 | 68.6 | 75.6 | | 58.6 | 59.2 | 52.7 | 62.8 | 71.4 | 86.8 | | 65.2 | 53.6 | 52.0 | 59.2 | 69.1 | 76.4 | AVG | 58.4 | 54.5 | 55.2 | 57.4 | 73.8 | 82.4 |
|
|