Fusion of Motif- and Spectrum-Related Features for Improved EEG-Based Emotion Recognition

Table 4

Comparison of different feature selection algorithms and number of features (nof) for combined benchmark-motif feature set.

Feature groups	Valence		Arousal
	BACC	nof	BACC	nof
ANOVA	0.5930	40	0.5446	39
mRMR	0.5816	29	0.5645	17
RFE	0.6010	38	0.5598	20