Research Article
Automatic Detection of Atrial Fibrillation from ECG Signal Using Hybrid Deep Learning Techniques
Table 4
Performance of ResNet and RBF models using five-fold cross-validation.
| | Validation number | Epochs | Accuracy (%) | F1 score (%) |
| | 1 | 5 | 63.11 | 73.93 | | 10 | 68.24 | | 15 | 73.67 | | 20 | 79.51 |
| | 2 | 5 | 62.04 | 78.36 | | 10 | 71.28 | | 15 | 77.34 | | 20 | 80.67 |
| | 3 | 5 | 63.86 | 79.98 | | 10 | 70.27 | | 15 | 78.37 | | 20 | 81.61 |
| | 4 | 5 | 61.99 | 82.80 | | 10 | 69.33 | | 15 | 76.38 | | 20 | 82.63 |
| | 5 | 5 | 64.79 | 85.96 | | 10 | 72.58 | | 15 | 80.67 | | 20 | 84.56 |
| | Overall F1 score (%) | 80.20 |
|
|