| | Authors (year) | Classifier | Feature extraction | Classes | Subsets | Accuracy (%) |
| | Tang and Durand [10] (2012) | SVM | Filter bank, Teager energy, power, Lempel–Ziv complexity | 2 | (A, D), E | 98.72 | | Song et al. [22] (2016) | Initial ELM | DWT, Mahalanobis distance, sample entropy | 2 | D, E | 97.53 | | Güler et al. [29] (2005) | ANN | Lyapunov exponents | 3 | A, D, E | 96.79 | | Liang et al. [30] (2010) | SVM | Principal component analysis, approximate entropy, power | 3 | A, D, E | 98.67 | | Murugavel and Ramakrishnan [8] (2016) | SVM | DWT, largest Lyapunov exponent, approximate entropy | 3 | A, D, E | 96 | | Riaz et al. [31] (2016) | SVM | Empirical mode decomposition, temporal, spectral features | 3 | A, D, E | 85 | | This work | SELM | LDWT, maximum, standard deviation | 3 | A, D, E | 98.4 |
|
|
