From: Intelligent diagnosis with Chinese electronic medical records based on convolutional neural networks
Depth | One-layer CNN(%) | Two-layer CNN(%) | Three-layer CNN(%) | ||||||
---|---|---|---|---|---|---|---|---|---|
Fold ∖metrics | Precision | Accuracy | F1-score | Precision | Accuracy | F1-score | Precision | Accuracy | F1-score |
1 | 84.26 | 84.1 | 84.16 | 83.13 | 82.9 | 82.97 | 83.05 | 82.74 | 82.84 |
2 | 83.63 | 83.5 | 83.52 | 82.65 | 82.42 | 82.5 | 82.32 | 81.53 | 81.66 |
3 | 83.86 | 83.55 | 83.61 | 82.54 | 82.26 | 82.35 | 79.09 | 78.89 | 78.94 |
4 | 84.07 | 83.75 | 83.84 | 82.78 | 82.51 | 82.58 | 82.28 | 82.02 | 82.05 |
5 | 83.87 | 83.71 | 83.76 | 82.97 | 82.81 | 82.85 | 82.6 | 82.37 | 82.4 |
Average | 83.94 | 83.72 | 83.78 | 82.81 | 82.58 | 82.65 | 81.87 | 81.51 | 81.58 |