Table 4.
Smart Models Used in Studies.
| Smart Model | Count |
|---|---|
| Convolutional neural network | 23 |
| Support vector machines | 20 |
| Decision tree | 10 |
| Long short-term memory | 10 |
| Random forest | 9 |
| K-nearest neighbors | 8 |
| Artificial neural networks | 5 |
| Naïve Bayes | 5 |
| Not identified | 5 |
| Logistic regression | 4 |
| Multilayer perceptron | 4 |
| Recurrent neural networks | 3 |
| Elastic net logistic model | 2 |
| Gradient boosting | 2 |
| Gradient boosting decision tree | 2 |
| Neural network | 2 |
| A custom model based on thresholding of Shannon entropy | 1 |
| Convolution–recurrent hybrid model (CRNN) | 1 |
| Deep neural network with a softmax regression model | 1 |
| Deep residual network (ResNet) | 1 |
| Enhanced version of recurrent neural network named ERNN | 1 |
| Gramian angular fields (GAFs) | 1 |
| Hidden Markov model | 1 |
| Hybrid decision model | 1 |
| Layered hidden Markov model | 1 |
| Linear regression | 1 |
| Mixed-kernel-based extreme learning machine (MKELM) | 1 |
| Ridge regression | 1 |
| Sequential covering algorithm | 1 |
| Shallow wavelet scattering network (ScatNet) | 1 |
| Time-synchronous averaging | 1 |
| Time-span convolutional neural network | 1 |