Algorithm 2 Cluster State Transition Matrix Generation Algorithm

Input:$C==\left\{c_1, c_2, \dots , c_K\right\}$ /* set of cluster centroids */

$K$ /* number of clusters */

$d_{max}$ /* distance limit */

Output:$P=\left\{p_{11}, \dots , p_{KK}\right\}$ /* set of cluster transition matrix */

for $i\leftarrow 1$ to $K$ do

$c_{min}\leftarrow i$;

$d_{min}\leftarrow 0$;

$flag\leftarrow false$;

for $j\leftarrow 1$ to $K$ do

$d\leftarrow \mathrm{distance}\left(c_i, c_j\right)$;

if $i=j$ or $d\le d_{max}$ then

$a_{ij}\leftarrow 1$;

$flag\leftarrow true$;

else

$a_{ij}\leftarrow 0$;

if $c_{min}=i$ or $d_{min}\ge d$ then

$c_{min}\leftarrow j$;

$d_{min}\leftarrow d$;

end

end

end

if $flag=false$ then

$j\leftarrow c_{min}$;

$a_{ij}\leftarrow 1$;

end

end

for $i\leftarrow 1$ to $K$ do

$near\_num\leftarrow {\sum }_{j=1}^K{a}_{ij}$;

for $j\leftarrow 1$ to $K$ do

if $a_{ij}=1$ then

$p_{ij}\leftarrow \frac{1}{near\_num}$;

else

$p_{ij}\leftarrow 0$;

end

end

end