Graph Based Multi-Layer K-Means++ (G-MLKM) for Sensory Pattern Analysis in Constrained Spaces

. 2021 Mar 16;21(6):2069. doi: 10.3390/s21062069

Algorithm 1. Subgraph Extraction

1:
Input: $\forall b_{i j} \in A^{*}$ ;
2:
Output: $(O_{T}^{i_{n t s}}, I_{T}^{i_{n t s}})$ , $i_{n t s} \in {a, b, c, \dots}$
3:
$I d x_{r o w}$ = $I d x_{c o l}$ = ${1, 2, \dots, | A^{*} |};$
4:
i = 0;
5:
for $i_{n t s}$ in ${a, b, c, \dots}$ do
6:
$O_{T}^{i_{n t s}} = I_{T}^{i_{n t s}} = \emptyset;$
7:
if $| I d x_{r o w} | \geq 1$ then
8:
procedure Increment(i)
9:
i = i + 1;
10:
if $i \in$ $I d x_{r o w}$ then
11:
return i;
12:
else
13:
Increment(i);
14:
procedure Recursion(i)
15:
if $\sum_{\forall j \in I d x_{c o l}} b_{i j} \geq 1$ then
16:
$O_{T}^{i_{n t s}} = O_{T}^{i_{n t s}} \cup \{i\};$
17:
procedure Extract(i)
18:
for j in $I d x_{c o l}$ do
19:
if $b_{i j} \neq 0$ then
20:
$I_{T}^{i_{n t s}} = I_{T}^{i_{n t s}} \cup \{j\};$
21:
$I d x_{c o l}$ = $I d x_{c o l}$ $\ I_{T}^{i_{n t s}};$
22:
$I d x_{r o w}$ = $I d x_{r o w}$ $\ {i};$
23:
for $j \in I_{T}^{i_{n t s}}$ do
24:
if $\sum_{\forall l \in I d x_{r o w}} b_{l j} \geq 1$ then
25:
for l in $I d x_{r o w}$ do
26:
if $b_{l j} \neq 0$ then
27:
$O_{T}^{i_{n t s}} = O_{T}^{i_{n t s}} \cup \{l\};$
28:
if $I d x_{r o w} \cap O_{T}^{i_{n t s}} \neq \emptyset$ then
29:
Extract $(\exists l \in (I d x_{r o w} \cap O_{T}^{i_{n t s}}))$ ;
30:
else
31:
return $(O_{T}^{i_{n t s}}, I_{T}^{i_{n t s}})$ ;
32:
else
33:
$I d x_{r o w}$ = $I d x_{r o w}$ $\ {i};$
34:
i = i + 1;
35:
Recursion $(i)$ ;
36:
else
37:
break;