Table 1 Summaries of methods of previous studies and their corresponding results.
References | FE methods | Classifier(s) | Dataset | Classification accuracy (%) |
|---|---|---|---|---|
– | 13 layer CNN | Malaysian dataset | 88.25 | |
– | CNN + LSTM | UNM dataset | 99.2 | |
– | CNN + RNN | Own dataset | 99.2 | |
Gabor transformation + 2D-CNN | SanDiego dataset | 99.44 | ||
Smoothed pseudo-Wigner Ville distribution + CNN | SanDiego dataset | 100 | ||
WT + Shanono entropy | No used | Finnish dataset | No score | |
WT + sample entropy | Three-way decision model | Chinese dataset | 92.68 | |
Higher-order spectra (HOS) | DT,KNN,FKNN, NB,PNN,SVM | Malaysian dataset | 90.6–99.6 | |
PSD | Hyperplanes | UNM dataset | 85.4 | |
WT + statistical measures | SVM | SanDiego dataset | 96.13 | |
CSP + entropy | SVM/KNN | SanDiego + UNM | 99 | |
