Skip to main content
. 2018 Aug 1;176:203–214. doi: 10.1016/j.neuroimage.2018.04.029

Table 3.

Motor and Imagery tasks classification.

Participant Number of Trials Evaluation CCR
Number of Selected Features
Mean SD Mean Minimum maximum
Motor Task 1 1050 0.73 0.01 29 21 35
2 1050 0.84 0.02 47 39 52
3 1050 0.76 0.02 39 37 42
4 1050 0.58 0.04 30 24 40
5 1050 0.66 0.02 34 29 37
6 1050 0.73 0.02 31 28 34
7 1050 0.40 0.02 23 18 29
8 1050 0.76 0.02 30 24 38
9 1050 0.65 0.03 37 32 48
10 750 0.64 0.03 25 16 31
11 1050 0.54 0.02 32 25 39
12 1050 0.71 0.03 39 31 49
13 1050 0.82 0.02 41 30 53
14 1050 0.83 0.03 30 16 39
15 1050 0.84 0.01 36 29 44
Imagery Task 1 1050 0.50 0.04 27 21 36
2 1050 0.63 0.03 29 18 38
3 900 0.68 0.04 32 21 39
4 1050 0.47 0.03 26 20 32
5 1050 0.35 0.02 22 15 33
6 1050 0.51 0.04 30 19 34
7 1050 0.30 0.05 18 12 25
8 1050 0.56 0.02 27 20 36
9 1050 0.61 0.02 30 25 34
10 1050 0.56 0.03 29 23 42
11 1050 0.38 0.02 13 9 16
12 1050 0.65 0.03 41 38 46
13 1050 0.84 0.02 29 23 33
14 1050 0.81 0.02 45 41 48
15 1050 0.69 0.02 15 10 17

The average correct classification rate (CCR) of the evaluation ERP data for the Motor and Imagery tasks. Classifiers were trained using randomly selected subsets (60% of the data) and this was repeated 5 times. The trained classifiers were applied on the unseen evaluation data (40% of the data). SD: is the standard deviation over the 5 repeats.