Research Article
Medical Text Classification Using Hybrid Deep Learning Models with Multihead Attention
Table 10
Analysis of results with different learning rates.
| Learning rate | Dataset | Accuracy (%) | Precision (%) | Recall (%) | F1-score (%) |
| 0.1 | Hallmarks | 63.85 | 56.16 | 55.81 | 55.98 | AIM | 88.17 | 82.36 | 81.28 | 81.81 |
| 0.01 | Hallmarks | 74.18 | 73.58 | 72.93 | 73.25 | AIM | 95.12 | 87.87 | 87.27 | 87.56 |
| 0.001 | Hallmarks | 74.41 | 69.91 | 69.98 | 69.94 | AIM | 82.64 | 77.09 | 76.91 | 76.99 |
| 0.0001 | Hallmarks | 70.71 | 64.71 | 65.91 | 65.30 | AIM | 61.36 | 55.16 | 57.73 | 56.41 |
|
|