Using Black Hole Algorithm to Improve EEG-Based Emotion Recognition

Algorithm 1

Black hole algorithm.

Require: number of stars (solutions) and stop criteria (maximum of iterations)

Ensure: the black hole. In our case, the best classifier.

produce the first generation of stars in the search space.

select the best solution as black hole.

while a good enough solution has not been reached in a maximum of iterations do

for all star do

evaluate classifier performance using cross-validation of data.

change the location of according to Eq.(5).

if is better than black hole then

select the current solution as black hole.

endif

cross to the event horizon defined by Eq.(6)

if then

replace with a new star in a random location in the search space.

endif

endfor

endwhile

return results and visualization.