Fuzzy -Means and Cluster Ensemble with Random Projection for Big Data Clustering

Algorithm 3

Cluster ensemble for FCM clustering with random projection.

Input: data set (an matrix), number of clusters , reduced dimension , number of

random projection , FCM clustering algorithm;

Output: cluster label vector .

() at each iteration , run Algorithm 2, get membership matrix ;

() concatenate the membership matrices ;

() compute the first left singular vectors of , denoted by ,

where , is a diagonal matrix and ;

() treat each row of as a data point and apply -means to obtain cluster label vector.