Fast Constrained Spectral Clustering and Cluster Ensemble with Random Projection

Algorithm 2

Spectral ensemble clustering with random projection.

Input: binary matrix , weights set , cluster number .

Output: the final partition result.

() Generate a sparse random matrix meeting the requirements of Lemma 2,

   where , , , ;

() Compute , where is a diagonal matrix with diagonal entries ;

() Compute ;

() Run weighted -means clustering on with weight set to obtain the final clustering result.