Table 3.

Arousal classification accuracy obtained using different feature extraction techniques on frequency bands.

Method	$\mathit{\delta }$	$\mathit{\theta }$	$\mathit{\alpha }$	$\mathit{\beta }$	$\mathit{\gamma }$	ALLBAND
DE	82.01%	88.10%	87.78%	77.96%	80.65%	88.47%
RASM	57.55%	58.06%	64.08%	76.34%	74.49%	59.42%
DCAU	71.96%	75.90%	75.35%	75.27%	74.52%	94.60%
DASM	75.13%	81.03%	79.64%	79.31%	81.06%	94.17%
AsMap+CNN	81.38%	88.27%	87.24%	88.94%	89.00%	95.21%

The window size was set to 3 s.