Research Article
Supervised Learning for Suicidal Ideation Detection in Online User Content
Table 6
Comparison of different models using all processed features on Twitter data.
| Model | Acc. | Prec. | Recall | F1 | AUC |
| Random Forest | 0.9638 | 0.9638 | 0.9917 | 0.9646 | 0.9862 | GBDT | 0.9500 | 0.9413 | 0.9603 | 0.9503 | 0.9825 | XGBoost | 0.9591 | 0.9425 | 0.9782 | 0.9597 | 0.9843 | SVM | 0.9485 | 0.9261 | 0.9755 | 0.9497 | 0.9813 | MLFFNN | 0.9412 | 0.9661 | 0.9194 | 0.9421 | 0.9823 | LSTM | 0.9108 | 0.9399 | 0.8802 | 0.9059 | 0.9747 |
|
|