Research Article
An Improved Double Channel Long Short-Term Memory Model for Medical Text Classification
Table 4
The results of the experiment.
| Model | Dataset | Accuracy | Precision | Recall | F1-score |
| CNN | cMedQA | 0.9570 | 0.8739 | 0.8803 | 0.8760 | Sentiment140 | 0.8831 | 0.8189 | 0.8188 | 0.8188 |
| LSTM | cMedQA | 0.9605 | 0.8857 | 0.8851 | 0.8852 | Sentiment140 | 0.9083 | 0.8255 | 0.8267 | 0.8256 |
| CNN-LSTM | cMedQA | 0.9615 | 0.8892 | 0.8898 | 0.8877 | Sentiment140 | 0.9010 | 0.8272 | 0.8271 | 0.8271 |
| GRU | cMedQA | 0.9611 | 0.8896 | 0.8795 | 0.8816 | Sentiment140 | 0.9044 | 0.8533 | 0.8532 | 0.8532 |
| DC-LSTM | cMedQA | 0.9729 | 0.9184 | 0.9183 | 0.9192 | Sentiment140 | 0.9112 | 0.8703 | 0.8703 | 0.8703 |
|
|