Research Article
Exploring Sampling in the Detection of Multicategory EEG Signals
Table 5
Comparison the results of our proposed approach with some reported research outcomes.
| Methods | Description | Classification accuracy | Overall performance | Class Z |
Class O | Class N | Class F | Class S |
| Proposed approach | RS + -NN | 100.0 | 100.0 | 100.0 | 100.0 | 100 | 100.0 | Siuly and Li [5] | Optimum allocation + MLS-SVM | 100.0 | 100.0 | 100.0 | 100.0 | 99.96 | 99.99 | Shen et al. [6] | Wavelet-approximate entropy + SVM | 100.0 | 100.0 | 99.87 | 100.0 | 100.0 | 99.97 |
Übeyli [9] | Wavelet coefficients and power spectral density (PSD) values + SVM | 99.25 | 99.13 | 99.25 | 99.38 | 99.00 | 99.20 | Übeyli [10] | PSD values + SVM | 99.38 | 99.25 | 99.13 | 99.50 | 99.25 | 99.30 | Guler and Ubeyli [11] | Wavelet coefficients and Lyapunov exponents + SVM | 99.25 | 99.38 | 99.25 | 99.38 | 99.13 | 99.28 |
|
|