Table 2.

Weighted average recall and standard deviation of speaker-dependent SER experiments with feature selection.

Dataset	MLP	SVM	RF	KNN
Emo-DB	$94.12\pm 2.31$	$95.10\pm 2.25$	$93.70\pm 2.68$	$90.80\pm 2.59$
SAVEE	$81.50\pm 4.20$	$82.10\pm 4.38$	$81.70\pm 4.52$	$78.31\pm 4.98$
IEMOCAP	$83.80\pm 3.14$	$69.10\pm 4.47$	$81.20\pm 3.08$	$76.66\pm 3.20$
RAVDESS	$79.20\pm 3.24$	$81.30\pm 3.21$	$80.60\pm 3.68$	$81.10\pm 3.95$