Novel Approaches to Identify Clusters Using Independent Components Analysis with Application

Algorithm 1

EMC_clustering_algorithm.

	Input: : matrix of observations, : number of clusters
	Output: C: clusters
(1)	;/^∗ execute ICA algorithm to generate ICs ^∗/
	/^∗compute estimated mixing matrix W and the corresponding demixing matrix A^∗/
(2)	;
(3)	;
	/^∗compute sum of squares of mixing coefficients in ^∗/
(4)	for to
(5)	for to
(6)
(7)	end for
(8)	end for
(9)	sortAscending(sumVect);
	/^∗ cluster the ordered rows in sumVect into k clusters, using an arbitrary clustering scheme ^∗/
(10)