Research Article
Fusion of Motif- and Spectrum-Related Features for Improved EEG-Based Emotion Recognition
Table 6
Top 20 features used in the best arousal models for the different feature groups.
| Benchmark (nof = 15, FS = RFE) | Motif (nof = 16, FS = RFE) | Combined (nof = 17, FS = mRMR) |
| | (β) | (θ) | | (β) | (γ) | AI(O1, O2) (β) | (β) | (θ) | | (α) | (α) | AI(P7, P8) (γ) | (α) | (β) | AI(F3, F4) (β) | (α) | (β) | | (θ) | (α) | Spectral power (β) | (γ) | (θ) | AI(Cp5, Cp6) (θ) | (γ) | (θ) | AI(FC1, FC2) (α) | (β) | (γ) | AI(P3, P4) (θ) | (β) | (θ) | AI(P3, P4) (β) | C (α) | (θ) | AI(C3, C4) (θ) | (β) | (α) | AI(Cp1, Cp2) (α) | (α) | (θ) | AI(T7, T8) (γ) | (α) | (α) | | k (α) | (α) | | | (γ) |
|
|