Table 4.
Comparison of performance of the proposed approach against other existing algorithms on the MIT-BIH arrhythmia database.
| Work | Approach | Accuracy (%) |
|---|---|---|
| ELP | CNN | 97.00 |
| ELP | RNN | 96.96 |
| ELP | RNN-Attention | 97.00 |
| Kachuee et al. [10] | Deep residual CNN | 93.4 |
| Acharya et al. [1] | Augmentation + CNN | 93.47 |
| Li et al. [13] | DWT + random forest | 94.61 |
| Martis et al. [17] | DWT + SVM | 93.8 |
DWT: Discrete wavelet transform; SVM: Support vector machine