Table 7.
Comparative results of deep learning with traditional machine-learning models.
Model | Recall | Precision | F1-score | Accuracy |
Logistic Regression | 0.42 | 0.34 | 0.36 | 0.42 |
AdaBoosta | 0.49 | 0.48 | 0.46 | 0.49 |
Gradient Boost | 0.50 | 0.59 | 0.45 | 0.50 |
ANNb | 0.29 | 0.08 | 0.13 | 0.29 |
kNNc | 0.35 | 0.36 | 0.37 | 0.35 |
Proposed Bi-LSTMd model | 0.93 | 0.94 | 0.94 | 0.93 |
aAdaBoost: adaptive boosting.
bANN: artificial neural network.
cKNN: k-nearest neighbor.
dBi-LSTM: bidirectional long-short term memory.