Fig. 3
From: A novel channel selection method for optimal classification in different motor imagery BCI paradigms
Overall accuracy curves showing the accuracy behaviors with the varying numbers of channels. For each subject, channel is added one by one in term of its channel weight. Channels with large weights have the priority of being added first. The overall accuracy curves are the average over eight subjects in each dataset. The red line is the mean accuracy curve and the blue dot lines represent the envelopes of ±standard deviation. a Average accuracy curve for MI tasks paradigm; b average accuracy curve for two-class control paradigm; c average accuracy curve for four-class control paradigm