Algorithm 1.

Constrained K-means

1:	Initialize K, n
2:	Input: Normalized data X, initial centroids $\left\{{\overline{x}}_1,\dots ,{\overline{x}}_{K+1}\right\}$ and cell representatives {R1, R2…..RK + 1}
3:	For iter = 1, 2, … , n
4:	Cluster assignment:
5:	$L_i=\{\begin{array}{l}k\text{when}{x}_i\in R_k\hfill \\ {\text{argmin}}_k{\Vert x_i-{\overline{x}}_k\Vert }^2,\text{otherwise}\hfill \end{array}$
6:	Centroid computation
7:	${\overline{x}}_k=\frac{1}{\left|C_k\right|}\sum _{x_i\in C_k}{x}_i$
8:	End For
9:	Return: Labels L, Centroids $\overline{x}$