Algorithm 1: Spectral Clustering Algorithm
	Input: Dataset, number of clusters k, parameter σ and number of iterations $iter$ Output: the set of k clusters /* Note the symbols of “$/*$” and “$*/$” represent comments in this algorithm. */
1	Calculate the affinity matrix $A\in {\mathbb{R}}^{n\times n}$ and define $A_{ij}=exp(-\parallel s_i-s_j{\parallel }^2/2{\sigma }^2)$ /* In which $S_i$ and $S_j$ are the original data points i and j, repctively. */
2	If $i\ne j$ then the $A_{ii}=0$
3	The D is the diagonal degree matrix and computed with elements: $d_i={\sum }_{j=1}^n{A}_{ij}$ Given a graph G with n input vertices, the Laplacian matrix $L(n\times n)=D-A$
4	Find the k largest eigenvectors of the matrix L and $[x_1{x}_2\cdots x_k]\in {\mathbb{R}}^{(n\times k)}$
5	Generate matrix y by renormalizing each x row, $Y_{ij}=X_{ij}/{({\sum }_j{X}_{ij}^2)}^{1/2}$
6	Minimize the distortion of each row Y to regard as the point in clustering term ${\mathbb{R}}^k$ using any clustering algorithm, such as a distance-based clustering approach.
7	Finally, the original point $s_i$ is assigned to cluster j when the row of $y_i$ belongs to the cluster j.
8	return the set of k clusters and cluster centre.