Research Article
Medical Text Classification Using Hybrid Deep Learning Models with Multihead Attention
Table 5
Analysis of results with different optimizers for the proposed quad channel hybrid model.
| Optimizer | Dataset | Accuracy (%) | Precision (%) | Recall (%) | F1-score (%) |
| Adam | Hallmarks | 72.98 | 69.65 | 71.61 | 70.61 | AIM | 95.12 | 87.17 | 85.99 | 86.57 |
| SGD | Hallmarks | 71.72 | 63.58 | 61.84 | 62.69 | AIM | 82.35 | 78.98 | 77.21 | 78.08 |
| Nadam | Hallmarks | 71.43 | 65.98 | 63.28 | 64.60 | AIM | 86.76 | 72.98 | 72.31 | 72.64 |
| AdaGrad | Hallmarks | 70.61 | 61.56 | 60.81 | 61.18 | AIM | 94.41 | 85.13 | 85.26 | 85.19 |
|
|