Table 3 The performance (weighted macro-average metric) of CNN based binary classification model on validation set
From: An application of convolutional neural networks with salient features for relation classification
Group No | Accuracy | Precision | Recall | F1 score |
|---|---|---|---|---|
| Â | Max/Ave/Min | Max/Ave/Min | Max/Ave/Min | Max/Ave/Min |
Group 1 | 89.81/77.19/57.27 | 72.92/56.85/45.22 | 51.85/50.65/49.62 | 94.63/86.26/71.14 |
Group 2 | 90.28/82.42/72.29 | 77.10/72.45/69.55 | 70.93/64.45/58.61 | 94.79/88.29/76.83 |
Group 3 | 90.97/81.45/69.05 | 80.40/72.36/66.74 | 66.90/62.11/58.32 | 95.18/87.90/75.64 |
Group 4 | 90.28/81.22/68.13 | 79.69/71.30/66.96 | 64.66/59.67/53.13 | 94.85/88.10/76.37 |
Group 5 | 89.58/78.4/59.35 | 87.35/65.23/57.90 | 60.08/53.81/50.92 | 94.48/86.94/72.59 |