Table 7.

Normalized confusion matrix of the proposed network with a symmetric parameter.

		Predicted Class
Actual class		(L = 1, S = 1)			(L = 2, S = 2)			(L = 4, S = 4)
		Normal	Drowsy		Normal	Drowsy		Normal	Drowsy
Normal		0.856	0.144		0.923	0.077		0.940	0.060
Drowsy		0.118	0.882		0.107	0.893		0.072	0.928
Actual class		(L = 6, S = 6)			(L = 8, S = 8)			(L = 10, S = 10)
		Normal	Drowsy		Normal	Drowsy		Normal	Drowsy
Normal		0.946	0.054		0.945	0.055		0.945	0.055
Drowsy		0.067	0.933		0.064	0.936		0.061	0.939