Discovering Human Activities from Binary Data in Smart Homes

. 2020 Apr 29;20(9):2513. doi: 10.3390/s20092513

Algorithm 8 computeSimilarity.

Input:

{\hat{\vec{P}}}_{i}^{s}, {\hat{\vec{c}}}_{j}^{s}, {\hat{\vec{P}}}_{j}^{s}, {\hat{\vec{T}}}_{i}^{s}, {\hat{\vec{T}}}_{j}^{s}, T_{s}, N + 2

Output: $\hat{\vec{S}}, \hat{\vec{Z}}$

1:
$\hat{r} = c o m p u t e N o r m C o n s t a n t (T_{s}, N + 2)$ //Algorithm 7
2:
for $h \leftarrow 1$ to $| {\hat{\vec{P}}}_{i}^{s} |$ do
3:
for $g \leftarrow 1$ to $| {\hat{\vec{P}}}_{j}^{s} |$ do
4:
${\hat{S}}_{(h, g)} = - 1$
5:
${\hat{Z}}_{(h, g)} = - 1$
6:
end for
7:
end for
8:
for $h \leftarrow 1$ to $| {\hat{\vec{P}}}_{i}^{s} |$ do
9:
for $g \leftarrow 1$ to $| {\hat{\vec{P}}}_{j}^{s} |$ do
10:
if ${\hat{t}}_{h}^{s} \neq {\hat{t}}_{g}^{s}$ then
11:
break
12:
end if
13:
$\hat{v} = c o m p u t e E M D ({\hat{\vec{p}}}_{h}^{s}, {\hat{\vec{p}}}_{g}^{s})$ //Algorithm 6
14:
$e m d = 1 - \frac{\hat{v}}{\hat{r}}$
15:
if $e m d > β$ then
16:
${\hat{S}}_{(h, g)} = e m d$
17:
${\hat{Z}}_{(h, g)} = {\hat{o}}_{g}^{s}$
18:
end if
19:
end for
20:
end for