Figure 7. Accuracy when inputs for a classifier were changed.

(A) Classification accuracy when the number of channels was reduced. The bold black, red, and blue lines indicate mean accuracy across subjects using 64 channels, 19 channels, and 6 channels, respectively. (B) Classification accuracy when different temporal features were used for classification. The data for 10-averaged trials is shown. Note that this analysis was done with 64 channels. Each colored line represents accuracy from an individual subject and the bold black line indicates the mean accuracy across subjects.