Table 10.
References | No. of classes | Feature set | Classifier | Accuracy |
---|---|---|---|---|
Li et al. (2014) | 5 | WPE + RR | RF | 94.61% |
Osowski et al. (2004) | 13 | HOS + Herminte | SVM | 98.18% |
Martis et al. (2013) | 5 | Cumulant + PCA | NN | 94.52% |
Acharya et al. (2017a) | 2 | End-to-end | CNN | 95.22% |
Acharya et al. (2017b) | 5 | End-to-end | CNN | 94.03% |
Zubair et al. (2016) | 5 | End-to-end | CNN | 92.70% |
Pławiak and Acharya (2020) | 17 | Frequency components | DGEC | 94.60% |
Atal and Singh (2020) | 5 | Gabor Filter + Wavelet | BaROA-DCNN | 93.19% |
The proposed CNN network | 5 | Wavalet | CNN | 97.20% |