| Method | Feature | Representative research | Recognition rate (%) |
| Linear analysis methods (Pearson correlation, amplitude squared coherence, autoregressive model, cumulative energy algorithm, time-frequency analysis, etc.) | EEG signal waveform characteristics (such as amplitude, phase, etc.), rhythm wave average power, power spectral density, band energy, wavelet coefficient root mean square, etc. | References [6] | 60.42 | References [7] | 62.50 | References [37] | 88.51 | Nonlinear analysis methods (mutual information [38], correlation dimension, Lempel–Ziv (LZ) complexity, recursive graph, and entropy analysis [39]) | Entropy, fractal dimension, correlation dimension, CO complexity, LZ complexity, Hust index, maximum Lyapunov index, etc. | References [40] | 80.40 | References [41] | 92.50 | References [42] | 86.65 |
|
|