Table 4 Data segementation with a tenfold scheme for a combination of three datasets (MIT-BIH Atrial Fibrillation, the 2017 PhysioNet/CinC Challenge, the China Physiological Signal Challenge 2018 databases)
From: AFibNet: an implementation of atrial fibrillation detection with convolutional neural network
Fold | Training data | Validation data | Total | ||
|---|---|---|---|---|---|
N | AF | N | AF | Â | |
1 | 16,485 | 32,149 | 1790 | 3614 | 54,038 |
2 | 16,391 | 32,243 | 1884 | 3520 | 54,038 |
3 | 16,424 | 32,210 | 1851 | 3553 | 54,038 |
4 | 16,402 | 32,232 | 1873 | 3531 | 54,038 |
5 | 16,469 | 32,165 | 1806 | 3598 | 54,038 |
6 | 16,520 | 32,114 | 1755 | 3649 | 54,038 |
7 | 16,476 | 32,158 | 1799 | 3605 | 54,038 |
8 | 16,453 | 32,181 | 1822 | 3582 | 54,038 |
9 | 16,416 | 32,219 | 1859 | 3544 | 54,038 |
10 | 16,439 | 32,196 | 1836 | 3567 | 54,038 |