Fast Enhanced Exemplar-Based Clustering for Incomplete EEG Signals

Algorithm 1

Fast enhanced exemplar-based clustering algorithm.

	Input: Given incomplete data
	Output: Valid exemplar set .
(1)	Compression: Apply basic exemplar-based clustering algorithm to the complete data ;
(2)	Get the potential exemplar set by equation (11) and construct the new similarity matrix ;
(3)	Generate the expansion order on the potential exemplar list
(4)	Let ;
(5)	for do
(6)	if then
(7)	compute by equations (15) and (16);
(8)	if then
(9)	for , set ;
(10)	end
(11)	else
(12)	compute by equations (17)–(19)
(13)	if then
(14)	for , set
(15)	else
(16)	for , set
(17)	else
(18)	if then
(19)	Accept the new exemplar
(20)	end
(21)	end
(22)	t = t + 1
(23)	end
(24)	Until converge.