A New Feature Analysis Approach to Selecting Channels of EEG for Fatigue Driving

Table 2

Recognition accuracy of each channel subset based on KNN classifier (Unit: %).

Number of channels	Theta_Std	Alpha_Std	Beta1_Std	Beta2_Std
1	77.22	54.90	68.53	58.60
2	90.37	67.92	82.90	69.37
3	94.73	78.67	90.88	72.92
4	96.35	84.23	93.03	78.33
5	96.63	86.38	94.07	79.92
6	99.37	87.85	94.63	82.65
7	99.35	88.60	95.95	82.60
8	99.28	89.77	95.78	83.65
9	99.33	89.98	96.25	84.32
10	99.33	89.85	96.30	84.80
11	99.28	90.72	96.43	85.45
12	99.37	90.90	96.60	85.87
13	99.30	91.07	96.82	85.20
14	99.32	91.02	96.83	86.53
15	99.42	91.10	96.72	85.72
16	99.35	91.45	96.63	85.92
17	99.35	91.28	97.00	86.13
18	99.33	91.12	96.73	85.55
19	99.22	91.73	96.92	85.92
20	99.33	91.57	96.78	86.70
21	99.30	91.57	96.88	85.92
22	99.27	91.37	96.88	86.08
23	99.27	91.70	96.80	85.97
24	99.20	91.40	96.80	86.10
25	99.22	91.52	96.90	85.92
26	99.22	91.55	96.97	85.87
27	99.23	91.57	96.83	85.92
28	99.22	91.27	96.83	85.73
29	99.25	91.65	96.82	85.87
30	99.23	91.38	96.98	85.97