Table 2.

The performance of different classification algorithms on 5-class-dataset according to GOS0: Acc, Rec, and Avg stand for accuracy, recall, and average, respectively.

Algorithm	Acc (%)	Acc Rank	Rec1 (%)	Rec2 (%)	Rec3 (%)	Rec4 (%)	Rec5 (%)
NB	44.76 ± 3.94	4	42.95	33.51	11.51	9.60	81.66
RF	45.37 ± 1.53	3	52.66	10.47	1.19	1.66	94.06
KNN(k = 5)	33.82 ± 2.07	8	34.48	9.42	8.33	17.55	60.61
KNN(k = 6)	36.36 ± 2.55	7	40.44	10.99	7.54	17.88	64.18
DT	43.86 ± 1.96	6	53.29	5.24	1.59	1.32	91.17
RI	44.71 ± 3.02	5	44.20	17.80	11.90	8.28	86.42
DL	46.22 ± 1.60	2	55.49	8.90	11.90	11.92	85.57
GBT	47.67 ± 2.65	1	59.25	27.23	8.73	10.60	83.70
Avg Acc	42.85