Table 3 Results of the different convolution kernel shapes and different numbers of convolution layers
From: Prediction of exercise sudden death in rabbit exhaustive swimming using deep neural network
| Different layers | Convolution kernel size | Acc (%) | Sens (%) | Spec (%) |
|---|---|---|---|---|
| 4 layers | 10 | 72.71 | 80.71 | 74.25 |
| 20 | 75.33 | 78.16 | 70.72 | |
| 30 | 65.77 | 72.15 | 51.38 | |
| 6 layers | 10 | 73.45 | 77.02 | 79.90 |
| 20 | 77.48 | 79.58 | 77.35 | |
| 30 | 75.03 | 86.97 | 73.05 | |
| 8 layers | 10 | 85.35 | 87.91 | 84.75 |
| 20 | 89.57 | 89.38 | 92.20 | |
| 30 | 85.72 | 88.29 | 87.42 |