Research Article
Fusion of Motif- and Spectrum-Related Features for Improved EEG-Based Emotion Recognition
Table 5
Top 20 features used in the best valence models for the different feature groups.
| Benchmark (nof = 3, FS = RFE) | Motif (nof = 20, FS = RFE) | Combined (nof = 38, FS = RFE) |
| | (α) | C (α) | | (γ) | | Spectral power (α) | (θ) | | | (θ) | | | (θ) | (θ) | | C (θ) | (γ) | | (β) | (β) | | S (β) | (θ) | | S (γ) | (θ) | | (γ) | (θ) | | (γ) | C (θ) | | (β) | Spectral power (α) | | (θ) | C (β) | | (β) | (β) | | (β) | (θ) | | (θ) | (θ) | | (θ) | (β) | | (θ) | (γ) | | (θ) | (θ) | | (β) | (γ) |
|
|