TABLE 3.

The comparative training results (with standard deviation) in three-class classification task.

Algorithms	Accuracy	Sensitivity	Specificity	Precision	F1-score	G-mean
LD	80.13	80.70	80.52	79.57	80.94	80.61
	(2.72)	(1.92)	(2.24)	(2.28)	(2.34)	(2.36)
LR	81.31	82.55	80.25	81.17	81.82	81.39
	(2.55)	(2.30)	(2.03)	(2.62)	(2.19)	(2.16)
SRC	81.94	82.20	80.46	81.09	81.17	81.33
	(2.20)	(2.49)	(2.59)	(2.10)	(2.21)	(2.54)
LC-KSVD	83.80	85.54	81.64	83.94	83.32	83.57
	(1.76)	(1.68)	(2.98)	(2.23)	(1.80)	(2.24)
FDDL	84.04	86.12	81.13	83.93	84.32	83.59
	(2.30)	(2.61)	(2.33)	(2.24)	(2.36)	(2.47)
SRDML	85.39	86.82	84.88	86.32	86.86	85.85
	(2.33)	(2.00)	(2.37)	(2.05)	(2.33)	(2.02)
LMLS-SRC	89.32	91.38	86.81	88.86	89.00	89.07
	(1.84)	(1.20)	(2.81)	(2.12)	(1.53)	(1.83)