Table 1 Model performance for different Batch-sizes when the dataset is divided in a ratio of 8:2.
From: An incremental adversarial training method enables timeliness and rapid new knowledge acquisition
Batch-size | Accuracy | Precision | Recall | F1-score |
|---|---|---|---|---|
8 | 0.9980 | 0.9976 | 0.9983 | 0.9979 |
16 | 0.9940 | 0.9945 | 0.9933 | 0.9938 |
32 | 0.9860 | 0.9855 | 0.9862 | 0.9858 |
64 | 0.9860 | 0.9867 | 0.9860 | 0.9863 |
128 | 0.9740 | 0.9733 | 0.9748 | 0.9738 |
