Table 2.
Studies regarding driver fatigue detection using different methods
| Research group | Feature method | Classifiers | EEG channels | Average accuracy (%) |
|---|---|---|---|---|
| Xiong et al. (2016) | AE and SE | SVM | P3 | 90.0 |
| Correa et al. (2014) | Multimodal Analysis | NN | C3-O1, C4-A1 and O2-A1 | 83.6 |
| Nguyen et al. (2017) | Statistical tests | Fisher’s linear discriminant analysis | All channels | 88.6 |
| Ko et al. (2015) | Fast Fourier Transformation | The linear regression | A single frontal channel | 90.0 |
| Chai (Chai et al. 2017) | AR | Sparse-DBN | All channels | 93.1 |
| This paper | FE | GBDT | TP7 | 94.0 |