Sleep Stage Classification Using Unsupervised Feature Learning

Table 1

Classification accuracy and 𝐹₁-score for the three experimental setups.

	Accuracy (mean ± std)	𝐹1-score
		Awake	S1	S2	SWS	REM
feat-GOHMM	63.9 ± 10.8	0.71	0.31	0.72	0.82	0.47
feat-DBN	72.2 ± 9.7	0.78	0.37	0.76	0.84	0.78
raw-DBN	67.4 ± 12.9	0.69	0.36	0.78	0.83	0.58