Table 3 The PACNN model with different embedding feature
From: Relation path feature embedding based convolutional neural network method for drug discovery
Methods | Precision(%) | Recall(%) | F-score(%) |
|---|---|---|---|
PACNN-random | 83.21 | 82.68 | 82.53 |
PACNN-one | 86.43 | 85.79 | 85.56 |
PACNN-pra | 91.50 | 91.50 | 91.46 |