I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
37
,
No
.
2
,
Feb
r
u
ar
y
20
25
,
p
p
.
758
~
770
I
SS
N:
2
502
-
4
7
52
,
DOI
: 1
0
.
1
1
5
9
1
/ijee
cs
.v
37
.
i
2
.
p
p
758
-
7
7
0
758
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
Na
tural sm
a
rt
ho
me a
utom
a
tion sy
stem using
LSTM
bas
ed on
ho
useho
ld behav
io
ur
M
o
cha
m
a
d Sus
a
nto
k
1,
2
,
F
a
r
ha
na
Ahm
a
d P
o
’
a
d
1
,
Arif
f
u
dd
in J
o
re
t
1
,
M
a
uli
na
H
ilwa
Sa
ls
a
bil
la
h
2
1
F
a
c
u
l
t
y
o
f
El
e
c
t
r
i
c
a
l
a
n
d
El
e
c
t
r
o
n
i
c
En
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
si
t
i
Tu
n
H
u
sse
i
n
O
n
n
M
a
l
a
y
s
i
a
,
P
a
r
i
t
R
a
j
a
,
M
a
l
a
y
s
i
a
2
P
r
o
g
r
a
m St
u
d
i
Te
k
n
i
k
E
l
e
k
t
r
o
n
i
k
a
T
e
l
e
k
o
mu
n
i
k
a
si
,
P
o
l
i
t
e
k
n
i
k
C
a
l
t
e
x
R
i
a
u
,
P
e
k
a
n
b
a
r
u
,
I
n
d
o
n
e
si
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l
16
,
2
0
2
4
R
ev
is
ed
Sep
18
,
2
0
2
4
Acc
ep
ted
Sep
30
,
2
0
2
4
A
sm
a
rt
h
o
m
e
a
u
to
m
a
ti
o
n
sy
ste
m
(S
HA
S
)
u
ti
li
z
in
g
d
a
ta
-
d
riv
e
n
lea
rn
in
g
i
s
a
n
a
d
v
a
n
c
e
d
in
tern
e
t
o
f
t
h
in
g
s
(
Io
T
)
a
p
p
li
c
a
ti
o
n
a
ime
d
to
lea
rn
h
o
u
se
h
o
l
d
b
e
h
a
v
i
o
r
to
p
re
v
e
n
t
m
in
iatu
r
c
ir
c
u
it
b
re
a
k
e
r
(M
CB)
tri
p
s
d
u
e
to
o
v
e
rlo
a
d
.
Un
li
k
e
trad
i
ti
o
n
a
l
d
e
term
in
isti
c
m
e
th
o
d
s,
t
h
is
stu
d
y
lev
e
ra
g
e
s
a
l
a
y
e
re
d
AI
m
o
d
e
l,
fe
a
tu
ri
n
g
re
a
l
-
ti
m
e
d
a
ta
c
o
ll
e
c
ti
o
n
,
l
o
n
g
sh
o
r
t
-
term
m
e
m
o
ry
(LS
T
M
)
b
a
se
d
lea
rn
i
n
g
,
a
n
d
a
n
a
u
to
m
a
ti
c
c
o
n
tro
l
sy
ste
m
.
Th
e
LS
T
M
c
l
a
s
sifica
ti
o
n
m
o
d
e
l
g
e
n
e
ra
tes
p
re
c
ise
ON
/OF
F
c
o
n
tro
l
sig
n
a
ls
se
n
t
to
Io
T
s
m
a
rtp
lu
g
s,
o
p
ti
m
izi
n
g
a
p
p
li
a
n
c
e
u
sa
g
e
a
n
d
r
e
d
u
c
in
g
th
e
ris
k
o
f
e
lec
tri
c
a
l
o
v
e
r
lo
a
d
.
Da
ta
fro
m
sm
a
rtp
lu
g
se
n
so
rs,
in
c
lu
d
in
g
a
p
p
li
a
n
c
e
sta
tu
s
a
n
d
e
n
v
ir
o
n
m
e
n
tal
fa
c
to
rs
li
k
e
p
o
we
r
c
o
n
su
m
p
ti
o
n
,
tem
p
e
ra
tu
re
,
a
n
d
h
u
m
id
it
y
,
we
re
c
o
ll
e
c
ted
e
v
e
ry
m
in
u
te
o
v
e
r
th
re
e
m
o
n
th
s
,
y
ield
in
g
8
0
,
8
1
8
d
a
ta
p
o
i
n
ts.
T
h
e
sy
ste
m
'
s
p
e
rfo
rm
a
n
c
e
wa
s
e
v
a
lu
a
ted
o
n
th
r
e
e
a
p
p
li
a
n
c
e
s:
Air
Co
n
d
it
i
o
n
e
r,
T
e
lev
isio
n
,
a
n
d
Wate
r
P
u
m
p
M
a
c
h
in
e
.
Re
su
l
ts
sh
o
we
d
h
i
g
h
a
c
c
u
ra
c
y
fo
r
Tel
e
v
isio
n
a
t
9
8
%
a
n
d
Wate
r
P
u
m
p
M
a
c
h
i
n
e
a
t
9
7
.
6
%
,
with
sli
g
h
t
ly
lo
we
r
a
c
c
u
ra
c
y
fo
r
Air
Co
n
d
it
io
n
e
r
a
t
8
1
.
9
%
.
T
h
is
d
e
m
o
n
stra
tes
th
e
sy
ste
m
'
s
e
ffe
c
ti
v
e
n
e
ss
in
re
a
l
-
wo
rld
a
p
p
li
c
a
ti
o
n
s.
T
h
e
sc
a
lab
il
it
y
a
n
d
a
d
a
p
ta
b
il
it
y
o
f
t
h
e
Na
tu
ra
l
S
HA
S
m
o
d
e
l
to
d
iffere
n
t
a
p
p
li
a
n
c
e
s
a
n
d
e
n
v
ir
o
n
m
e
n
ts
m
a
rk
a
sig
n
ifi
c
a
n
t
a
d
v
a
n
c
e
m
e
n
t
in
sm
a
rt
h
o
m
e
a
u
t
o
m
a
ti
o
n
,
o
ffe
ri
n
g
a
p
ra
c
ti
c
a
l
so
l
u
ti
o
n
f
o
r
p
re
v
e
n
ti
n
g
e
lec
tri
c
a
l
o
v
e
rl
o
a
d
a
n
d
imp
r
o
v
in
g
h
o
u
se
h
o
l
d
e
n
e
rg
y
m
a
n
a
g
e
m
e
n
t.
K
ey
w
o
r
d
s
:
E
lectr
ical
o
v
er
lo
ad
p
r
ev
e
n
tio
n
Ho
u
s
eh
o
ld
en
e
r
g
y
m
an
ag
e
m
e
n
t
I
o
T
s
m
ar
tp
lu
g
s
L
STM
R
ea
l
-
tim
e
d
ata
co
llectio
n
Sm
ar
t h
o
m
e
au
t
o
m
atio
n
s
y
s
te
m
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Far
h
an
a
Ah
m
ad
Po
’
ad
Facu
lty
E
lectr
ical
an
d
E
lectr
o
n
ic
E
n
g
in
ee
r
in
g
,
U
n
iv
er
s
iti T
u
n
Hu
s
s
ein
On
n
Ma
lay
s
ia
Par
it R
aja
,
J
o
h
o
r
,
8
6
4
0
0
,
Ma
la
y
s
ia
E
m
ail: f
ar
h
an
a@
u
th
m
.
e
d
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
Ho
u
s
eh
o
ld
s
in
I
n
d
o
n
esia
ar
e
c
ateg
o
r
ized
in
to
s
ev
er
al
cu
s
to
m
er
g
r
o
u
p
s
b
y
Per
u
s
ah
aa
n
L
is
tr
ik
Neg
ar
a
(
PLN)
,
th
e
s
tate
o
wn
ed
elec
tr
icity
co
m
p
an
y
,
in
ter
m
s
o
f
elec
tr
icity
co
n
s
u
m
p
tio
n
[
1
]
.
Fo
r
ex
am
p
le,
th
e
R
-
1
/TR
ca
teg
o
r
y
lim
it
elec
tr
icity
c
o
n
s
u
m
p
tio
n
to
a
m
a
x
im
u
m
o
f
1
,
3
0
0
VA
with
a
tar
if
f
o
f
R
p
.
1
,
4
4
4
.
7
0
p
e
r
k
W
h
.
T
h
is
m
ea
n
s
th
at
th
e
to
tal
el
ec
tr
icity
co
n
s
u
m
p
tio
n
f
r
o
m
all
ap
p
lian
ce
s
s
h
o
u
ld
n
o
t
ex
ce
ed
1
,
3
0
0
watts,
ass
u
m
in
g
a
p
o
wer
f
ac
to
r
(
c
o
s
p
h
i)
o
f
1
.
Ho
wev
er
,
in
r
e
ality
,
m
an
y
h
o
u
s
eh
o
ld
s
h
a
v
e
v
ar
io
u
s
elec
tr
ical
ap
p
lian
ce
s
with
a
to
tal
p
o
wer
co
n
s
u
m
p
tio
n
ex
ce
e
d
in
g
t
h
is
s
u
b
s
cr
ip
tio
n
lim
it.
C
o
n
s
eq
u
en
tly
,
wh
en
m
u
ltip
le
ap
p
lian
ce
s
ar
e
u
s
ed
s
im
u
ltan
eo
u
s
ly
,
ex
ce
e
d
in
g
th
e
to
tal
p
o
wer
lim
it,
it
ca
u
s
e
th
e
m
in
i
atu
r
e
cir
cu
it
b
r
ea
k
er
(
MCB
)
to
tr
ip
.
W
h
en
th
e
MC
B
tr
ip
s
,
th
e
elec
tr
ical
co
n
n
ec
ti
o
n
f
r
o
m
th
e
m
ain
PLN
lin
e
to
th
e
h
o
u
s
e
is
cu
t
o
f
f
,
ca
u
s
in
g
all
elec
tr
ical
ap
p
lian
c
es
in
th
e
h
o
u
s
e
to
s
h
u
t
d
o
wn
.
T
h
is
co
n
d
itio
n
lead
s
t
o
th
e
o
cc
u
r
r
en
ce
o
f
in
r
u
s
h
cu
r
r
en
t
i
n
elec
tr
ical
a
p
p
lian
c
es
wh
en
th
e
MCB
is
s
witch
ed
b
ac
k
o
n
.
T
h
e
in
r
u
s
h
c
u
r
r
en
t
p
h
e
n
o
m
en
o
n
is
d
is
ad
v
an
tag
eo
u
s
as
it
in
c
r
ea
s
es
elec
tr
icity
co
n
s
u
m
p
tio
n
at
th
e
b
eg
in
n
i
n
g
o
f
o
p
e
r
atio
n
d
u
e
to
th
e
lar
g
e
m
o
m
e
n
t
a
r
y
c
u
r
r
e
n
t
a
n
d
c
a
n
als
o
d
a
m
a
g
e
e
le
c
t
r
ic
a
l
e
q
u
i
p
m
e
n
t
b
e
c
a
u
s
e
o
f
t
h
e
la
r
g
e
s
t
a
r
t
i
n
g
t
o
r
q
u
e
r
e
q
u
i
r
e
d
[
2
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
N
a
tu
r
a
l sma
r
t h
o
me
a
u
to
ma
tio
n
s
ystem
u
s
in
g
LS
TM
b
a
s
ed
o
n
h
o
u
s
eh
o
ld
b
eh
a
vi
o
u
r
(
Mo
c
h
a
ma
d
S
u
s
a
n
to
k
)
759
Ad
d
itio
n
ally
,
f
r
eq
u
e
n
t
o
v
er
lo
ad
in
g
a
n
d
s
u
b
s
eq
u
en
t
tr
i
p
p
in
g
o
f
th
e
MCB
p
o
s
e
s
af
ety
r
is
k
s
to
th
e
h
o
u
s
eh
o
ld
[
3
]
,
as e
lectr
ical
eq
u
ip
m
en
t m
ay
b
e
d
am
a
g
ed
[
4
]
,
p
o
ten
tially
lead
in
g
to
d
an
g
er
o
u
s
s
itu
atio
n
s
[
5
]
.
Fu
r
th
er
m
o
r
e,
th
is
f
r
eq
u
en
t
ON
-
OFF
cy
clin
g
o
f
elec
tr
ic
ity
ca
n
d
is
r
u
p
t
en
er
g
y
m
a
n
ag
em
e
n
t
in
th
e
h
o
m
e,
ca
u
s
in
g
in
ef
f
icien
cies a
n
d
in
c
o
n
v
e
n
ien
ce
[
6
]
.
T
o
ad
d
r
ess
th
is
is
s
u
e,
au
to
m
atic
co
n
tr
o
l
s
y
s
tem
s
b
ased
o
n
lo
ad
p
r
io
r
itizatio
n
[
3
]
an
d
th
e
elim
in
atio
n
o
f
h
ig
h
-
co
n
s
u
m
p
tio
n
ap
p
lian
c
es
[
7
]
h
av
e
b
ee
n
im
p
lem
en
te
d
.
Ho
w
ev
er
,
th
ese
ap
p
r
o
ac
h
es
ar
e
o
f
ten
s
tatic
an
d
lack
ad
ap
ta
b
ilit
y
to
n
atu
r
al
e
n
v
ir
o
n
m
en
tal
f
ac
t
o
r
s
.
I
n
c
o
n
t
r
ast,
h
o
m
e
o
wn
er
s
wh
o
ar
e
a
war
e
o
f
th
eir
to
tal
p
o
wer
co
n
s
u
m
p
tio
n
ty
p
ically
d
ev
elo
p
h
a
b
its
o
f
tu
r
n
in
g
ap
p
l
ian
ce
s
ON
an
d
OF
F a
s
n
ee
d
ed
to
av
o
id
ex
ce
ed
in
g
th
e
m
ax
im
u
m
lo
ad
lim
it.
T
h
is
b
eh
av
io
r
r
ef
lects
a
m
o
r
e
d
y
n
am
ic
an
d
r
esp
o
n
s
iv
e
a
p
p
r
o
ac
h
to
en
er
g
y
m
an
ag
em
en
t
with
in
th
e
h
o
m
e
.
T
h
is
s
tu
d
y
f
o
cu
s
es
o
n
th
e
b
eh
av
io
r
o
f
h
o
u
s
eh
o
l
d
r
esid
en
t
s
as
th
e
o
b
ject
o
f
r
esear
ch
to
p
r
o
v
i
d
e
s
o
lu
tio
n
s
f
o
r
p
r
e
v
en
tin
g
o
v
e
r
lo
ad
s
.
T
h
e
m
eth
o
d
u
s
es
an
elec
tr
ical
ap
p
r
o
ac
h
b
y
lev
er
a
g
in
g
in
ter
n
et
o
f
th
in
g
s
(
I
o
T
)
tech
n
o
lo
g
y
to
ca
p
tu
r
e
d
ata
o
n
r
esid
e
n
ts
'
b
eh
av
io
r
an
d
AI
d
ee
p
lear
n
in
g
tech
n
o
lo
g
y
t
o
au
to
m
atica
lly
co
n
tr
o
l th
e
ON
-
OFF s
tate
o
f
elec
tr
ical
ap
p
lian
ce
s
b
ased
o
n
th
e
r
esid
en
ts
'
b
eh
av
io
r
.
T
h
e
in
teg
r
atio
n
o
f
AI
an
d
I
o
T
k
n
o
wn
as
AI
o
T
h
as
r
esu
lted
in
im
p
r
ess
iv
e
tech
n
o
lo
g
y
w
h
er
e
d
ee
p
lear
n
in
g
ca
n
d
er
iv
e
n
ew
in
s
ig
h
ts
f
r
o
m
co
n
tin
u
o
u
s
ly
u
p
d
ated
d
ata
f
r
o
m
v
a
r
io
u
s
s
o
u
r
ce
s
.
T
h
is
ad
v
an
ce
m
en
t h
as
b
r
o
u
g
h
t
p
r
o
g
r
ess
to
v
ar
io
u
s
f
i
eld
s
,
s
u
ch
as
s
m
ar
t
h
o
m
e
s
[
8
]
,
s
m
ar
t
cities
[
9
]
,
in
d
u
s
tr
y
,
h
e
alth
ca
r
e
[
1
0
]
,
[
1
1
]
,
an
d
tr
an
s
p
o
r
tatio
n
[
1
2
]
.
I
n
t
h
e
co
n
tex
t
o
f
s
m
ar
t
h
o
m
es,
th
e
ap
p
licatio
n
o
f
th
is
tech
n
o
lo
g
y
ex
ten
d
s
to
en
h
an
cin
g
th
e
f
u
n
ctio
n
ality
a
n
d
co
m
f
o
r
t
o
f
h
o
u
s
eh
o
l
d
r
esid
en
ts
.
T
h
is
in
clu
d
es
th
e
ev
o
l
u
tio
n
f
r
o
m
r
em
o
te
ac
ce
s
s
an
d
co
n
tr
o
l
to
au
to
m
a
tic
s
y
s
tem
co
n
tr
o
l
b
ased
o
n
lear
n
in
g
d
ata
[
1
3
]
,
th
er
e
b
y
im
p
r
o
v
in
g
r
esid
en
ts
'
co
m
f
o
r
t i
n
u
s
in
g
elec
tr
ical
ap
p
lian
ce
s
wh
ile
ac
h
iev
in
g
en
e
r
g
y
s
av
in
g
s
[
1
4
]
.
R
ef
er
r
in
g
to
I
o
T
tech
n
o
lo
g
y
,
elec
tr
ical
ap
p
lian
ce
s
with
s
m
ar
t
s
o
ck
e
ts
ca
n
tr
an
s
m
it
d
ata
o
n
th
e
s
tatu
s
o
f
an
ap
p
lian
ce
,
in
clu
d
i
n
g
p
o
wer
co
n
s
u
m
p
tio
n
,
c
u
r
r
en
t,
v
o
l
tag
e,
en
er
g
y
,
p
o
wer
f
ac
to
r
[
7
]
,
[
1
5
]
,
an
d
ON/OFF
s
tatu
s
[
1
6
]
to
o
th
er
d
ev
ices
a
n
d
th
e
clo
u
d
with
in
an
I
o
T
n
et
wo
r
k
.
Similar
ly
,
o
th
er
s
m
ar
t
h
o
m
e
d
e
v
ices
s
u
ch
as
th
er
m
o
s
tats
,
PIR
s
en
s
o
r
s
f
o
r
d
etec
tin
g
h
u
m
an
p
r
esen
ce
in
a
r
o
o
m
,
an
d
wate
r
tan
k
lev
el
s
en
s
o
r
s
to
tr
ig
g
er
t
h
e
wate
r
p
u
m
p
ca
n
s
er
v
e
as
d
ata
s
o
u
r
ce
s
f
o
r
s
m
ar
t
s
y
s
tem
s
lik
e
th
e
s
m
ar
t
h
o
m
e
au
t
o
m
atio
n
s
y
s
tem
(
SHAS)
to
u
n
d
er
s
tan
d
h
o
u
s
eh
o
ld
b
e
h
av
i
o
r
p
atter
n
s
in
o
p
er
at
in
g
elec
tr
ical
ap
p
lian
ce
s
.
Fu
r
th
er
m
o
r
e,
as
to
em
b
ed
d
in
g
AI
tech
n
o
lo
g
y
in
t
h
e
I
o
T
s
y
s
tem
,
Dee
p
lear
n
in
g
m
o
d
els
s
u
ch
as
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
ca
n
b
e
im
p
lem
en
ted
as
it
s
u
itab
le
f
o
r
lear
n
in
g
d
atasets
with
tim
e
s
er
ies
d
ata,
lik
e
h
o
u
s
eh
o
ld
b
eh
av
io
r
p
atte
r
n
s
th
at
f
o
llo
w
d
aily
c
y
cles
[
1
7
]
-
[
1
8
]
.
On
ce
p
atter
n
s
ar
e
id
en
tifie
d
th
r
o
u
g
h
d
ee
p
lear
n
i
n
g
,
o
r
wh
at
is
r
ef
er
r
ed
to
as
th
e
m
o
d
el,
SHAS
will
u
s
e
th
is
m
o
d
el
to
p
r
ed
ict
o
u
tp
u
ts
in
th
e
f
o
r
m
o
f
au
to
m
atic
co
n
tr
o
ls
,
s
u
ch
as
tu
r
n
i
n
g
a
n
elec
tr
ical
ap
p
lian
ce
ON
o
r
OF
F.
T
h
u
s
,
au
to
m
atic
co
n
tr
o
l
h
a
p
p
en
s
n
atu
r
ally
r
ath
e
r
th
a
n
d
et
er
m
in
is
tically
[
1
9
]
,
th
r
o
u
g
h
co
n
d
itio
n
in
g
o
r
s
ch
ed
u
lin
g
,
s
u
ch
as
th
e
s
ch
ed
u
le
d
ON/OFF
co
n
tr
o
l
o
f
lig
h
ts
[
2
0
]
.
T
h
is
ad
d
s
v
al
u
e
to
SHAS with
th
e
ap
p
licatio
n
o
f
AI
in
an
I
o
T
en
v
ir
o
n
m
en
t.
B
ased
o
n
an
AI
o
T
s
y
s
tem
,
th
i
s
s
tu
d
y
co
n
tr
ib
u
tes
to
th
e
d
esig
n
o
f
a
lay
er
ed
SHAS
m
o
d
el
u
s
in
g
th
e
L
STM
alg
o
r
ith
m
with
d
atasets
f
r
o
m
I
o
T
d
ev
ices
in
a
h
o
u
s
e,
in
clu
d
in
g
th
e
ON
-
OFF
s
tatu
s
o
f
elec
tr
ical
ap
p
lian
ce
s
an
d
o
t
h
er
s
u
p
p
o
r
tin
g
s
en
s
o
r
d
ata.
T
h
e
f
o
llo
win
g
s
ec
tio
n
s
o
f
t
h
is
p
ap
er
will
co
v
er
th
e
m
eth
o
d
o
lo
g
y
o
f
th
e
b
u
ilt
s
y
s
tem
an
d
its
wo
r
k
f
lo
w,
f
o
llo
wed
b
y
a
d
is
cu
s
s
io
n
o
f
th
e
r
esu
lts
,
wh
ich
will
p
r
esen
t
th
e
d
ataset
ch
ar
ac
ter
is
tics
an
d
th
e
p
e
r
f
o
r
m
an
ce
ev
alu
atio
n
o
f
L
STM
in
p
r
ed
ictin
g
th
e
ON
o
r
OFF
s
tatu
s
o
f
elec
tr
ical
ap
p
li
an
ce
s
u
s
in
g
th
e
c
o
n
f
u
s
io
n
m
atr
ix
p
a
r
am
eter
.
2.
M
E
T
H
O
D
T
h
is
s
tu
d
y
p
r
o
p
o
s
es
a
SHAS
lay
er
ed
m
o
d
el,
as
d
ep
icted
in
Fig
u
r
e
1
,
wh
ich
in
teg
r
ates
AI
with
in
an
I
o
T
e
n
v
ir
o
n
m
en
t.
T
h
e
SHAS
d
ev
elo
p
e
d
in
th
is
r
esear
ch
u
til
izes
L
STM
as
th
e
AI
f
ea
tu
r
e
t
o
co
n
tr
o
l
elec
tr
ical
ap
p
lian
ce
s
b
ased
o
n
h
o
u
s
eh
o
ld
b
eh
av
i
o
r
d
ata.
I
n
th
e
p
h
y
s
ical
lay
er
,
s
en
s
o
r
s
g
en
er
ate
d
ata
s
u
ch
as
th
e
ON/OFF
s
tatu
s
o
f
s
m
ar
tp
lu
g
s
o
r
s
m
ar
t
s
o
ck
ets
f
o
r
elec
tr
ical
ap
p
lian
ce
s
,
to
tal
elec
tr
icity
co
n
s
u
m
p
tio
n
f
r
o
m
s
m
ar
t
p
o
wer
m
eter
s
,
air
q
u
ality
d
ata
in
clu
d
in
g
tem
p
e
r
atu
r
e
a
n
d
h
u
m
id
ity
,
wate
r
tan
k
lev
el
d
ata
in
d
i
ca
tin
g
th
e
p
er
ce
n
tag
e
o
f
wate
r
h
eig
h
t
in
th
e
tan
k
,
an
d
m
o
tio
n
d
ata
d
et
ec
tin
g
h
u
m
an
p
r
esen
ce
in
a
r
o
o
m
.
T
h
ese
s
en
s
o
r
s
wir
eless
ly
tr
an
s
m
it
th
eir
d
ata
with
in
an
I
o
T
n
etwo
r
k
to
a
n
ed
g
e
d
ev
ice,
f
u
n
ctio
n
i
n
g
a
s
a
s
en
s
o
r
h
u
b
th
at
ag
g
r
eg
ates
all
s
en
s
o
r
d
ata
in
t
o
a
s
tr
ea
m
in
g
d
ataset
n
am
ed
th
e
au
to
m
atio
n
h
o
m
e
elec
tr
ical
ap
p
lian
ce
co
n
tr
o
l
s
y
s
tem
(
AHE
AC
S).
T
h
e
d
ataset
d
u
r
atio
n
is
p
er
m
in
u
te,
c
o
llected
o
v
er
t
h
r
ee
m
o
n
th
s
f
r
o
m
Feb
r
u
ar
y
t
o
Ap
r
il
2
0
2
4
,
to
talin
g
8
0
,
8
1
8
d
ata
p
o
i
n
ts
.
I
n
th
e
m
a
n
ag
em
e
n
t
lay
er
,
L
STM
p
r
o
ce
s
s
es
th
e
d
ataset
to
l
ea
r
n
,
ev
al
u
ate
th
e
m
o
d
el'
s
p
er
f
o
r
m
an
ce
,
an
d
p
r
o
d
u
ce
th
e
m
o
s
t
o
p
tim
al
p
r
ed
ictiv
e
m
o
d
el
to
b
e
u
s
ed
in
t
h
e
p
r
o
p
o
s
ed
SHAS
m
o
d
el.
T
h
e
m
o
d
el'
s
p
r
ed
ictio
n
r
esu
lts
,
in
t
h
e
f
o
r
m
o
f
ON/OFF
co
m
m
an
d
s
,
a
r
e
s
en
t
to
th
e
s
en
s
o
r
d
e
v
ice
s
,
s
p
ec
if
ically
th
e
s
m
ar
tp
lu
g
s
co
n
tain
in
g
r
elay
s
co
n
n
ec
ted
to
th
e
elec
tr
ical
ap
p
lian
ce
s
o
ck
ets.
T
h
is
co
m
m
u
n
icatio
n
ca
n
b
e
ca
r
r
ied
o
u
t
th
r
o
u
g
h
a
lo
ca
l
I
o
T
n
etwo
r
k
s
u
ch
as
wir
eless
lo
ca
l
ar
ea
n
etwo
r
k
(
W
L
AN)
an
d
b
y
u
s
in
g
m
ac
h
in
e
-
to
-
m
ac
h
in
e
co
m
m
u
n
icatio
n
p
r
o
to
co
ls
lik
e
m
ess
ag
e
q
u
eu
i
n
g
telem
etr
y
t
r
an
s
p
o
r
t
(
MQ
T
T
)
.
I
n
th
e
ca
s
e
o
f
co
m
m
u
n
icatio
n
b
etwe
en
I
o
T
d
ev
ices,
s
u
ch
as
AI
d
ev
ices
an
d
s
m
ar
tp
lu
g
s
,
th
ey
ac
t
as
M
QT
T
clien
ts
with
an
MQ
T
T
b
r
o
k
er
d
e
v
ice
th
at
is
al
s
o
co
n
n
ec
ted
lo
ca
lly
with
in
th
e
s
am
e
W
L
A
N.
T
h
is
AI
p
r
o
ce
s
s
ca
n
b
e
ex
ec
u
ted
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
a
r
y
20
25
:
7
5
8
-
7
7
0
760
at
th
e
e
d
g
e,
en
s
u
r
i
n
g
th
at
t
h
e
au
t
o
m
atic
co
n
tr
o
l
p
r
o
ce
s
s
d
o
es
n
o
t
r
ely
o
n
in
ter
n
et
co
n
n
ec
tiv
ity
[
2
1
]
.
Me
an
wh
ile,
th
e
ap
p
licatio
n
la
y
er
r
esid
es
in
th
e
clo
u
d
,
p
r
o
v
id
in
g
u
s
er
s
with
th
e
o
p
tio
n
t
o
m
o
n
ito
r
s
en
s
o
r
d
ata
an
d
m
an
u
ally
r
em
o
te
-
co
n
t
r
o
l
elec
tr
ical
ap
p
lian
ce
s
v
ia
th
e
in
ter
n
et.
T
h
is
wo
r
k
p
r
esen
ts
th
e
p
er
f
o
r
m
an
ce
o
f
L
STM
as
th
e
AI
co
m
p
o
n
e
n
t
in
th
e
I
o
T
SHAS
ap
p
licatio
n
o
n
th
e
ed
g
e
s
id
e
o
f
an
I
o
T
n
etwo
r
k
,
as
s
h
o
wn
i
n
Fig
u
r
e
1
.
T
h
e
m
eth
o
d
o
l
o
g
y
co
n
s
is
ts
o
f
th
r
ee
p
ar
ts
:
th
e
s
tu
d
y
o
f
p
r
o
p
o
s
ed
d
ataset
c
h
ar
ac
ter
is
tics
,
th
e
d
esig
n
o
f
th
e
L
STM
m
o
d
el
f
o
r
ea
c
h
s
m
ar
t
p
lu
g
co
n
n
e
cted
to
a
n
elec
tr
ical
ap
p
lian
ce
an
d
also
p
er
f
o
r
m
a
n
c
e
an
aly
s
is
o
f
th
e
L
STM
alg
o
r
ith
m
in
Go
o
g
le
C
o
lab
en
v
ir
o
n
m
en
t u
s
in
g
p
y
th
o
n
.
Fig
u
r
e
1
.
SHAS lay
er
ed
m
o
d
e
l
2
.
1
.
SH
AS da
t
a
s
et
s
t
re
a
m
in
g
T
h
e
r
o
le
o
f
th
e
s
en
s
o
r
h
u
b
in
th
e
SHAS
lay
er
ed
m
o
d
el
d
esig
n
o
f
th
is
r
esear
c
h
s
h
o
wed
in
Fig
u
r
e
1
is
to
f
o
r
m
a
s
tr
ea
m
in
g
AHE
AC
S
d
ataset
f
r
o
m
v
ar
io
u
s
s
en
s
o
r
s
in
a
s
m
ar
t
h
o
m
e
ap
p
licatio
n
.
T
h
is
s
tu
d
y
u
s
es
a
s
m
ar
t
p
o
wer
m
eter
to
g
en
e
r
ate
to
tal
p
o
wer
(
T
P)
d
ata,
i
n
d
icatin
g
to
tal
elec
tr
icity
co
n
s
u
m
p
tio
n
.
Fiv
e
s
m
ar
tp
lu
g
s
p
r
o
v
id
e
ON/OFF
s
tatu
s
d
ata
f
o
r
ea
ch
elec
tr
ical
ap
p
lian
ce
,
s
u
c
h
as
an
air
co
n
d
itio
n
er
(S
-
AC
)
,
R
ef
r
ig
er
ato
r
(
S
-
K)
,
W
ater
Pu
m
p
(
S
-
W
PM)
,
W
ash
in
g
Ma
ch
in
e
(
S
-
MC),
an
d
T
elev
is
io
n
(
S
-
T
V)
.
Ad
d
itio
n
ally
,
en
v
ir
o
n
m
en
tal
s
en
s
o
r
s
th
at
r
elate
d
to
th
e
h
o
m
e'
s
co
n
d
itio
n
ar
e
u
s
ed
as
f
ac
t
o
r
s
in
f
lu
e
n
cin
g
th
e
r
esid
en
ts
'
d
ec
is
io
n
to
tu
r
n
elec
tr
ical
a
p
p
lian
ce
s
ON
o
r
OFF.
T
h
ese
in
clu
d
e
s
en
s
o
r
s
f
o
r
tem
p
e
r
atu
r
e
(
T
)
,
h
u
m
id
ity
(
H)
,
m
o
tio
n
in
th
e
b
e
d
r
o
o
m
(
Pb
)
,
m
o
tio
n
in
th
e
liv
in
g
r
o
o
m
(
P
g
)
wh
er
e
th
e
telev
is
io
n
is
lo
c
ated
,
an
d
wate
r
tan
k
lev
el
(
W
)
,
in
d
icatin
g
th
e
wate
r
lev
el
in
th
e
wate
r
tan
k
o
f
th
e
ch
o
s
en
h
o
u
s
e.
T
ab
le
1
s
h
o
ws
th
e
d
esig
n
o
f
t
h
e
s
tr
ea
m
in
g
d
ataset
co
lu
m
n
s
,
wh
er
e
th
e
tim
estam
p
(
T
S)
r
ep
r
esen
ts
th
e
d
ata
co
llectio
n
tim
e,
s
et
p
er
m
in
u
te.
S
-
X
r
ep
r
esen
ts
s
m
ar
t
p
lu
g
s
tatu
s
wh
ic
h
co
n
tain
S
-
AC
,
S
-
K,
S
-
W
PM,
S
-
MC,
an
d
S
-
T
V,
with
th
e
p
o
s
s
ib
ilit
y
o
f
a
d
d
in
g
o
th
e
r
elec
tr
i
ca
l
ap
p
lian
ce
s
in
th
e
f
u
tu
r
e.
I
n
th
is
s
tu
d
y
,
T
S
is
d
etailed
to
r
ef
lect
th
e
in
f
l
u
en
ce
o
f
wee
k
d
a
y
s
(
Mo
n
d
a
y
-
Fri
d
ay
)
an
d
wee
k
en
d
s
(
Satu
r
d
ay
-
Su
n
d
ay
)
,
as
well
as
o
p
er
atio
n
al
h
o
u
r
s
(
m
o
r
n
in
g
-
a
f
ter
n
o
o
n
-
ev
en
i
n
g
)
.
T
h
u
s
,
T
S
i
s
ex
tr
ac
ted
in
to
y
ea
r
,
m
o
n
th
,
d
ay
,
h
o
u
r
,
m
in
u
te,
an
d
s
ec
o
n
d
v
alu
es,
wh
ich
ar
e
co
n
v
er
ted
i
n
to
n
u
m
er
ic
f
o
r
m
.
Yea
r
an
d
m
o
n
th
r
em
ain
n
u
m
e
r
ic,
wh
ile
th
e
d
ay
is
r
ep
r
esen
ted
as
(
D)
,
wh
er
e
M
o
n
d
ay
=
1
,
T
u
esd
a
y
=2
,
an
d
s
o
o
n
u
n
til
Su
n
d
ay
=7
.
H
o
u
r
s
,
m
in
u
tes,
an
d
s
ec
o
n
d
s
ar
e
co
n
v
er
ted
i
n
to
th
e
n
u
m
b
e
r
o
f
s
ec
o
n
d
s
s
in
ce
m
id
n
ig
h
t
(
NSM)
,
a
v
alu
e
th
at
r
ep
ea
ts
d
aily
.
T
o
d
is
tin
g
u
is
h
b
etwe
en
wee
k
d
ay
s
an
d
wee
k
e
n
d
s
,
th
e
v
ar
ia
b
le
(
W
S)
is
u
s
ed
: 0
f
o
r
wee
k
d
ay
s
a
n
d
1
f
o
r
wee
k
en
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
N
a
tu
r
a
l sma
r
t h
o
me
a
u
to
ma
tio
n
s
ystem
u
s
in
g
LS
TM
b
a
s
ed
o
n
h
o
u
s
eh
o
ld
b
eh
a
vi
o
u
r
(
Mo
c
h
a
ma
d
S
u
s
a
n
to
k
)
761
T
ab
le
1
.
Data
s
et
lab
elin
g
V
a
r
i
a
b
l
e
D
e
scri
p
t
i
o
n
Ty
p
e
o
f
d
a
t
a
TS
Ti
mes
t
a
m
p
D
a
t
e
a
n
d
Ti
m
e
N
S
M
N
u
mb
e
r
o
f
S
e
c
o
n
d
s
si
n
c
e
M
i
d
n
i
g
h
t
,
c
o
n
v
e
r
t
a
l
l
t
i
me
i
n
a
d
a
y
t
o
b
e
c
o
me
n
u
mb
e
r
o
f
se
c
o
n
d
N
u
meri
c
a
l
WS
W
e
e
k
d
a
y
st
a
t
u
s
B
o
o
l
e
a
n
(
w
e
e
k
d
a
y
=
0
,
w
e
e
k
e
n
d
=
1
)
D
Th
e
D
a
y
f
r
o
m
M
o
n
d
a
y
t
o
S
u
n
d
a
y
N
u
meri
c
a
l
(
M
o
n
d
a
y
=
1
u
n
t
i
l
S
u
n
d
a
y
=
7
)
HR
D
a
t
e
D
a
t
e
WK
Ti
me
Ti
me
TP
To
t
a
l
P
o
w
e
r
(
W
a
t
t
)
N
u
meri
c
a
l
S
-
X
Th
e
st
a
t
e
o
f
e
l
e
c
t
r
i
c
a
l
a
p
p
l
i
a
n
c
e
o
p
e
r
a
t
i
o
n
B
o
o
l
e
a
n
(
o
n
=
1
,
o
f
f
=
0
)
T
Te
mp
e
r
a
t
u
r
e
(
°
C
)
N
u
meri
c
a
l
H
H
u
mi
d
i
t
y
(
%)
N
u
meri
c
a
l
Pb
P
r
e
sen
t
p
e
o
p
l
e
i
n
b
e
d
r
o
o
m
B
o
o
l
e
a
n
(
p
r
e
se
n
t
=
1
,
e
mp
t
y
=
0
)
Pg
P
r
e
sen
t
p
e
o
p
l
e
i
n
t
e
l
e
v
i
s
i
o
n
r
o
o
m
B
o
o
l
e
a
n
(
p
r
e
se
n
t
=
1
,
e
mp
t
y
=
0
)
W
W
a
t
e
r
t
a
n
k
l
e
v
e
l
N
u
meri
c
a
l
2
.
2
.
L
ST
M
ba
s
ed
f
o
r
SH
AS
T
h
e
r
o
le
o
f
I
o
T
tech
n
o
lo
g
y
is
h
ig
h
ly
b
en
ef
icial
i
n
c
o
llectin
g
r
ea
l
-
tim
e
u
s
ag
e
d
ata
o
f
elec
tr
ical
ap
p
lian
ce
s
in
h
o
m
es,
wh
ich
ca
n
b
e
co
m
p
iled
i
n
to
a
d
atas
et.
T
h
is
d
ata
is
th
e
n
p
r
o
ce
s
s
ed
an
d
a
n
aly
ze
d
to
u
n
d
er
s
tan
d
u
s
er
u
s
ag
e
h
ab
its
,
lead
in
g
to
ac
tio
n
s
b
ased
o
n
n
ew
in
s
ig
h
ts
o
b
tain
ed
,
as
d
em
o
n
s
tr
ated
in
th
e
SHAS a
p
p
licatio
n
.
T
h
is
co
n
ce
p
t is k
n
o
wn
as th
e
in
te
r
n
et
o
f
b
eh
av
io
r
(
I
o
B
)
[
2
2
]
.
I
n
th
e
co
n
tex
t
o
f
SHAS,
wh
ich
f
o
c
u
s
es
o
n
th
e
tim
e
-
d
e
p
en
d
en
t
u
s
ag
e
b
eh
a
v
io
r
o
f
h
o
u
s
eh
o
ld
elec
tr
ical
ap
p
lian
ce
s
,
a
r
ec
u
r
r
en
t n
eu
r
al
n
etwo
r
k
(
R
NN)
,
an
e
x
ten
s
io
n
o
f
a
Neu
r
al
Netwo
r
k
m
o
d
el
s
p
ec
if
ically
f
o
r
lear
n
in
g
s
eq
u
en
tial
d
ata
o
v
er
tim
e,
ca
n
b
e
a
p
p
lied
[
2
3
]
.
T
h
e
ex
p
ec
tatio
n
f
r
o
m
a
n
R
NN
is
to
ca
p
tu
r
e
l
o
n
g
-
ter
m
d
ep
e
n
d
en
cies
s
o
th
at
all
p
ast
in
p
u
ts
ca
n
in
f
lu
e
n
ce
th
e
o
u
tp
u
t.
Ho
wev
er
,
R
NNs
f
ac
e
ch
allen
g
es
with
in
p
u
ts
th
at
ar
e
t
o
o
d
is
tan
t
in
t
h
e
p
ast.
Fo
r
tu
n
ately
,
th
ese
p
r
o
b
lem
s
ca
n
b
e
a
d
d
r
ess
ed
b
y
ap
p
ly
in
g
a
n
ev
o
lu
tio
n
o
f
th
e
R
NN
k
n
o
wn
as
th
e
L
S
T
M
m
o
d
el
[
2
4
]
.
T
h
e
L
STM
m
o
d
el
in
clu
d
es
s
p
ec
ialized
u
n
its
ca
lled
m
em
o
r
y
ce
lls
,
wh
ich
ca
n
s
to
r
e
in
f
o
r
m
atio
n
f
o
r
an
e
x
ten
d
ed
p
er
io
d
,
ac
tin
g
lik
e
a
c
o
n
v
e
y
o
r
c
o
n
n
ec
tin
g
L
STM
b
lo
ck
s
.
T
h
ese
u
n
its
in
v
o
lv
e
th
r
ee
ty
p
es
o
f
g
ates:
in
p
u
t,
f
o
r
g
et,
an
d
o
u
tp
u
t,
wh
ich
co
n
tr
o
l
th
e
f
lo
w
o
f
in
f
o
r
m
atio
n
.
T
h
ese
g
ates
ar
e
cr
u
cial
as
th
e
y
d
eter
m
in
e
wh
et
h
er
to
allo
w
n
ew
in
p
u
t,
er
ase
th
e
c
u
r
r
e
n
t
ce
ll
s
tatu
s
,
o
r
let
th
e
s
tatu
s
in
f
lu
en
ce
th
e
o
u
t
p
u
t a
t a
s
p
ec
if
ic
tim
e
s
tep
.
T
h
is
p
ap
er
d
o
es
n
o
t
ex
p
lain
t
h
e
g
en
er
al
wo
r
k
in
g
s
o
f
L
ST
M
in
d
etail,
as
L
STM
its
elf
is
n
o
t
n
ew
in
Dee
p
L
ea
r
n
in
g
alg
o
r
ith
m
s
.
F
o
r
m
o
r
e
d
etails,
r
ef
e
r
to
p
r
ev
io
u
s
liter
atu
r
e
o
n
L
STM
r
ela
ted
to
s
m
ar
t
h
o
m
e
ap
p
licatio
n
s
[
1
8
]
,
[
2
5
]
.
Her
e,
th
e
f
o
cu
s
is
o
n
th
e
d
etailed
L
STM
m
o
d
el
f
o
r
th
e
s
tr
ea
m
i
n
g
d
ataset
ca
s
e
as
p
r
ev
io
u
s
ly
e
x
p
lain
ed
.
T
h
e
L
S
T
M
ar
ch
itectu
r
e
in
th
is
s
tu
d
y
,
as
s
h
o
wn
in
Fig
u
r
e
2
,
co
n
s
is
ts
o
f
o
n
e
in
p
u
t
la
y
er
,
two
h
id
d
en
la
y
er
s
,
an
d
o
n
e
o
u
t
p
u
t la
y
er
with
f
u
lly
co
n
n
ec
ted
lay
er
.
Fig
u
r
e
2
.
L
STM
ar
c
h
itectu
r
e
f
o
r
SHAS
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
a
r
y
20
25
:
7
5
8
-
7
7
0
762
I
n
th
e
in
p
u
t
lay
er
o
f
t
h
e
L
ST
M
m
o
d
el,
th
e
n
u
m
b
er
o
f
in
p
u
t
f
ea
tu
r
es
is
d
eter
m
in
ed
b
y
v
ar
i
o
u
s
f
ac
to
r
s
th
at
in
f
lu
en
ce
th
e
ON/OFF
o
p
er
atio
n
o
f
ea
ch
elec
tr
ical
ap
p
l
ian
ce
.
Fo
r
ex
am
p
le,
in
th
e
ca
s
e
o
f
co
n
tr
o
llin
g
a
n
Air
C
o
n
d
itio
n
er
(
AC
)
,
r
elev
a
n
t
in
p
u
t
f
ea
tu
r
es
m
ig
h
t
in
clu
d
e
T
o
tal
Po
wer
(
T
P),
T
em
p
e
r
a
tu
r
e
(
T
)
,
Hu
m
id
ity
(
H)
,
B
ed
r
o
o
m
Mo
tio
n
(
Pb
)
,
an
d
W
ater
Pu
m
p
Statu
s
(
S
-
W
P
M)
.
T
h
u
s
,
f
o
r
th
e
s
m
ar
tp
lu
g
c
o
n
n
ec
ted
to
th
e
AC
,
th
e
L
STM
m
o
d
el
r
ec
eiv
es
5
in
p
u
t
f
ea
tu
r
es.
Ho
wev
er
,
with
in
th
e
L
STM
n
eu
r
al
n
etwo
r
k
,
th
e
in
p
u
t
lay
e
r
is
s
tr
u
ctu
r
ed
to
tak
e
in
o
n
e
tim
estep
o
r
in
p
u
t d
im
e
n
s
io
n
at
a
tim
e.
T
h
is
is
b
ec
au
s
e
all
in
p
u
t f
ea
tu
r
es a
r
e
m
ea
s
u
r
ed
s
im
u
ltan
eo
u
s
ly
at
a
s
p
ec
if
ic
tim
e
o
r
tim
estep
.
T
h
e
i
n
p
u
t
d
at
a,
r
ep
r
esen
ted
as
1
,
2
,
3
,
…
.
.
,
in
th
e
d
iag
r
a
m
,
co
r
r
esp
o
n
d
s
to
th
ese
f
ea
tu
r
es f
o
r
ea
ch
elec
tr
ical
ap
p
lian
ce
.
I
n
t
h
i
s
s
t
u
d
y
,
t
h
e
i
n
p
u
t
f
e
a
t
u
r
es
f
o
r
d
i
f
f
e
r
e
n
t
a
p
p
l
i
a
n
c
e
s
v
a
r
y
b
a
s
e
d
o
n
t
h
e
i
r
o
p
e
r
a
t
i
o
n
a
l
c
o
n
t
e
x
t
.
T
h
e
n
o
t
a
t
i
o
n
S
-
X
i
n
t
h
e
T
a
b
l
e
2
o
th
e
s
p
e
c
i
f
i
c
s
m
a
r
t
p
l
u
g
s
c
o
n
n
e
c
te
d
t
o
d
i
f
f
e
r
e
n
t
a
p
p
l
i
a
n
c
es
,
s
u
c
h
a
s
A
i
r
C
o
n
d
i
ti
o
n
e
r
(
S
-
AC
)
,
W
at
e
r
P
u
m
p
(
S
-
W
PM
)
,
R
e
f
r
i
g
e
r
at
o
r
(
S
-
K
)
,
W
a
s
h
i
n
g
M
a
c
h
i
n
e
(
S
-
MC
)
,
a
n
d
S
-
T
V
(
T
e
l
e
v
is
i
o
n
)
.
T
h
e
d
i
a
g
r
a
m
i
l
l
u
s
t
r
a
t
es
h
o
w
t
h
e
s
e
i
n
p
u
t
s
f
e
e
d
i
n
t
o
t
h
e
L
S
T
M
c
el
l
s
s
e
q
u
e
n
t
i
al
l
y
,
w
i
t
h
e
a
c
h
c
el
l
l
e
a
r
n
i
n
g
f
r
o
m
t
h
e
i
n
p
u
t
d
a
t
a
a
n
d
p
a
s
s
i
n
g
t
h
e
l
e
ar
n
e
d
i
n
f
o
r
m
a
t
i
o
n
f
o
r
w
a
r
d
t
h
r
o
u
g
h
t
h
e
n
e
t
w
o
r
k
,
u
l
t
i
m
at
e
l
y
l
ea
d
i
n
g
t
o
t
h
e
b
i
n
a
r
y
c
l
a
s
s
i
f
ic
a
t
i
o
n
o
u
t
p
u
t
t
h
a
t
c
o
n
t
r
o
l
s
t
h
e
a
p
p
li
a
n
c
e'
s
ON
/
OFF
s
t
a
t
e
.
T
h
is
p
r
o
c
es
s
i
s
c
a
p
tu
r
e
d
i
n
t
h
e
L
S
T
M
a
r
c
h
i
t
e
c
t
u
r
e
s
h
o
w
n
i
n
F
i
g
u
r
e
2
,
w
h
e
r
e
t
h
e
i
n
p
u
t
l
a
y
e
r
f
ee
d
s
t
h
e
d
at
a
i
n
t
o
a
s
e
r
i
es
o
f
L
S
T
M
c
e
l
ls
,
e
ac
h
m
a
i
n
t
a
i
n
i
n
g
a
n
d
u
p
d
a
t
i
n
g
t
h
e
c
e
l
l
s
t
a
t
e
(
)
a
n
d
h
i
d
d
e
n
s
t
a
t
e
ℎ
(
)
.
T
h
e
f
i
n
a
l
h
i
d
d
e
n
s
t
a
te
a
f
t
e
r
p
r
o
c
e
s
s
i
n
g
a
ll
t
i
m
e
s
t
e
p
s
i
s
t
h
e
n
u
s
e
d
f
o
r
b
i
n
a
r
y
c
l
a
s
s
i
f
i
ca
t
i
o
n
,
w
h
ic
h
d
e
t
e
r
m
in
e
s
t
h
e
o
p
e
r
a
t
i
o
n
al
c
o
n
t
r
o
l
s
t
a
tu
s
o
f
t
h
e
a
p
p
l
i
a
n
c
e
.
T
h
e
d
eter
m
in
at
io
n
o
f
in
p
u
t
f
ea
tu
r
es
is
b
ased
o
n
th
e
th
r
esh
o
ld
v
alu
e
o
f
t
h
e
co
r
r
elatio
n
co
ef
f
icien
t
with
th
e
(
1
)
:
=
∑
(
−
̅
)
(
−
̅
)
.
)
(
1
)
T
h
e
Pear
s
o
n
c
o
r
r
elatio
n
co
ef
f
i
cien
t
r
an
g
es
f
r
o
m
-
1
to
1
.
I
f
is
p
o
s
itiv
e,
th
e
in
p
u
t
f
ea
tu
r
e
v
ar
iab
les
ten
d
to
m
o
v
e
in
th
e
s
am
e
d
ir
ec
tio
n
,
in
d
icatin
g
a
p
o
s
itiv
e
co
r
r
elati
o
n
.
C
o
n
v
e
r
s
ely
,
if
is
n
eg
ativ
e,
th
e
co
r
r
elatio
n
is
n
eg
ativ
e,
m
ea
n
in
g
t
h
e
v
ar
iab
les
m
o
v
e
in
o
p
p
o
s
ite
d
ir
e
ctio
n
s
.
I
f
ap
p
r
o
ac
h
es
0
,
t
h
e
r
e
is
n
o
lin
ier
r
elatio
n
s
h
ip
b
etwe
en
th
e
two
i
n
p
u
t f
ea
tu
r
e
v
ar
iab
les.
Me
an
w
h
ile
r
ep
r
esen
t th
e
to
tal
n
u
m
b
e
r
o
f
d
ata
s
am
p
les
an
d
r
ep
r
esen
t th
e
s
tan
d
ar
d
d
e
v
iatio
n
f
r
o
m
two
v
a
r
iab
le.
T
h
e
co
r
r
elatio
n
lev
el
b
etwe
e
n
in
p
u
t
f
ea
tu
r
es
is
d
etailed
in
th
e
ex
p
lo
r
ato
r
y
d
ata
an
aly
s
is
(
E
DA
)
s
ec
tio
n
o
f
th
is
p
ap
e
r
.
Fro
m
th
e
in
p
u
t
la
y
er
,
a
f
u
lly
c
o
n
n
e
cted
n
etwo
r
k
is
cr
ea
te
d
to
th
e
f
ir
s
t
h
id
d
en
lay
er
,
wh
ich
co
n
s
is
ts
o
f
2
5
L
STM
c
ells
.
I
n
th
is
lay
er
,
th
e
g
en
er
al
L
STM
p
r
o
ce
s
s
o
cc
u
r
s
,
wh
er
e
m
em
o
r
y
ce
lls
with
th
r
ee
g
ates
ar
e
d
esig
n
ed
to
r
e
ad
,
s
to
r
e,
an
d
u
p
d
ate
p
ast
in
f
o
r
m
atio
n
.
E
ac
h
L
STM
ce
ll
co
n
n
ec
ts
to
th
e
n
ex
t
L
STM
ce
ll th
r
o
u
g
h
th
e
ce
ll st
ate
(
)
an
d
h
i
d
d
en
s
tate
ℎ
(
)
with
th
e
(
2
)
an
d
(
3
)
:
(
)
=
(
)
.
(
−
1
)
+
(
)
.
̃
(
)
(
2
)
ℎ
(
)
=
(
)
.
ta
n
h
(
(
)
)
(
3
)
W
h
er
e
is
:
(
)
=
f
o
r
g
et
g
ate
(
)
=
in
p
u
t g
ate
(
)
=
o
u
tp
u
t
g
ate
̃
(
)
=
ca
n
d
id
ate
g
ate
T
h
e
s
elec
tio
n
o
f
2
5
L
STM
ce
lls
,
as
s
h
o
wn
in
Fig
u
r
e
2
,
is
b
ased
o
n
th
e
co
m
p
le
x
ity
o
f
th
e
in
p
u
t
f
ea
tu
r
e
v
ar
iab
les
a
n
d
t
h
e
d
im
en
s
io
n
s
o
f
th
e
d
ata
[
2
6
]
.
As
p
r
ev
io
u
s
ly
m
en
tio
n
e
d
,
t
h
is
s
tu
d
y
em
p
l
o
y
s
a
s
in
g
le
in
p
u
t
d
im
en
s
io
n
f
o
r
tim
e
-
s
er
ies
d
ata,
aim
in
g
to
k
ee
p
th
e
m
o
d
el
s
im
p
le
an
d
s
u
itab
le
f
o
r
im
p
lem
en
tatio
n
o
n
th
e
ed
g
e
s
id
e
o
f
an
I
o
T
n
etwo
r
k
.
T
h
e
L
STM
n
etwo
r
k
th
en
co
n
n
ec
ts
to
th
e
s
ec
o
n
d
h
id
d
en
lay
e
r
,
wh
ich
m
ir
r
o
r
s
th
e
fi
r
s
t
lay
er
in
ter
m
s
o
f
p
ar
a
m
e
ter
s
,
s
u
ch
as
th
e
u
s
e
o
f
a
d
r
o
p
o
u
t
lay
er
a
n
d
th
e
r
et
u
r
n
s
eq
u
en
ce
s
ettin
g
.
T
h
e
d
r
o
p
o
u
t
lay
er
is
ap
p
lied
to
p
r
e
v
en
t
o
v
e
r
f
itti
n
g
d
u
r
in
g
tr
ain
i
n
g
,
with
a
d
r
o
p
o
u
t
r
ate
o
f
0
.
1
o
r
(
1
0
%)
o
f
t
h
e
to
tal
2
5
n
eu
r
o
n
s
p
er
h
id
d
en
lay
er
.
T
h
e
r
et
u
r
n
s
eq
u
e
n
ce
p
ar
am
ete
r
in
th
e
f
ir
s
t
h
id
d
en
lay
er
is
s
et
to
"tr
u
e"
to
en
s
u
r
e
th
at
th
e
o
u
tp
u
t
s
eq
u
en
ce
at
ea
ch
tim
estep
alig
n
s
with
th
e
in
p
u
t
co
n
d
itio
n
s
.
Fo
r
t
h
e
s
ec
o
n
d
h
id
d
en
lay
er
,
th
is
p
ar
am
eter
is
s
et
to
"f
alse"
s
o
t
h
at
th
e
o
u
tp
u
t
d
im
en
s
io
n
b
ec
o
m
es
s
in
g
u
lar
,
m
atch
in
g
th
e
d
esire
d
f
in
al
o
u
tp
u
t.
I
n
th
e
o
u
tp
u
t
la
y
er
,
th
e
L
STM
m
o
d
el
p
e
r
f
o
r
m
s
b
in
ar
y
class
if
icatio
n
,
p
r
ed
ictin
g
wh
eth
er
ea
ch
s
m
ar
tp
lu
g
co
n
n
ec
ted
t
o
an
elec
tr
ical
ap
p
l
ian
ce
s
h
o
u
ld
b
e
tu
r
n
ed
ON
(
1
)
o
r
OFF (
0
)
.
2
.
3
.
SH
AS
env
iro
nm
ent
T
h
is
s
tu
d
y
u
s
es
Go
o
g
le
C
o
lab
to
r
u
n
th
e
L
STM
alg
o
r
ith
m
with
Py
th
o
n
p
r
o
g
r
am
m
in
g
to
m
ea
s
u
r
e
its
p
er
f
o
r
m
an
ce
i
n
SHAS.
T
h
e
T
en
s
o
r
Flo
w
m
o
d
u
le
s
er
v
es
a
s
th
e
f
r
am
ewo
r
k
f
o
r
tr
ain
in
g
th
e
d
ee
p
lear
n
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
N
a
tu
r
a
l sma
r
t h
o
me
a
u
to
ma
tio
n
s
ystem
u
s
in
g
LS
TM
b
a
s
ed
o
n
h
o
u
s
eh
o
ld
b
eh
a
vi
o
u
r
(
Mo
c
h
a
ma
d
S
u
s
a
n
to
k
)
763
m
o
d
el,
with
Ker
as
as
th
e
in
teg
r
ated
n
eu
r
al
n
etwo
r
k
API
.
Ad
d
itio
n
ally
,
th
e
P
an
d
as
lib
r
ar
y
i
s
u
s
ed
f
o
r
th
e
E
DA
p
r
o
ce
s
s
to
d
eter
m
in
e
th
e
n
u
m
b
er
o
f
f
ea
tu
r
es
in
th
e
in
p
u
t
la
y
er
f
o
r
ea
ch
s
m
ar
tp
lu
g
elec
tr
i
ca
l
ap
p
lian
ce
o
f
t
h
e
d
ev
elo
p
e
d
SHAS
in
th
is
s
tu
d
y
.
Data
is
im
p
o
r
ted
f
r
o
m
th
e
s
en
s
o
r
h
u
b
in
.
csv
f
o
r
m
at
a
n
d
co
llected
wee
k
ly
,
with
d
im
en
s
io
n
s
o
f
(
8
0
,
8
1
8
;
1
2
)
,
r
e
p
r
esen
tin
g
8
0
,
8
1
8
r
o
ws
o
f
d
ata
a
n
d
1
2
c
o
lu
m
n
s
o
r
in
p
u
t
f
ea
tu
r
es.
T
h
e
T
S
co
lu
m
n
is
th
en
ex
tr
ac
ted
in
t
o
NSM,
W
S,
D,
HR
,
an
d
W
K,
as
s
h
o
wn
in
T
ab
le
1
,
ch
an
g
i
n
g
th
e
d
im
en
s
io
n
s
to
(
8
0
,
8
1
8
; 1
6
)
.
I
n
th
is
s
y
s
tem
,
wh
ich
u
s
es
an
AI
tech
n
o
lo
g
y
,
th
e
L
STM
tr
ain
in
g
d
ata
is
tak
en
wee
k
ly
,
ass
u
m
in
g
th
e
h
o
u
s
eh
o
l
d
r
esid
en
ts
'
h
ab
its
r
ep
ea
t
ea
ch
wee
k
o
n
n
o
r
m
al
d
ay
s
an
d
n
o
t
d
u
r
i
n
g
lo
n
g
h
o
li
d
ay
s
.
T
h
e
d
ataset
is
th
en
n
o
r
m
alize
d
to
s
p
ee
d
u
p
c
o
m
p
u
tatio
n
a
n
d
av
o
id
lar
g
e
v
alu
e
r
an
g
es
b
etwe
en
in
p
u
t
f
ea
tu
r
es
[1
3
]
u
s
in
g
t
h
e
s
cik
it
-
lear
n
lib
r
ar
y
in
Go
o
g
le
C
o
lab
.
T
h
is
n
o
r
m
aliza
tio
n
p
r
o
ce
s
s
in
v
o
lv
es
r
em
o
v
in
g
d
ata
with
NaN
o
r
n
u
ll
v
alu
es
o
r
ig
in
atin
g
f
r
o
m
th
e
s
e
n
s
o
r
d
ata
s
o
u
r
ce
s
.
T
h
e
o
cc
u
r
r
en
ce
o
f
s
u
ch
v
alu
es
d
ep
e
n
d
s
o
n
th
e
q
u
ality
o
f
th
e
s
en
s
o
r
r
ea
d
in
g
s
a
n
d
th
e
d
ata
n
etwo
r
k
d
u
r
in
g
tr
an
s
m
is
s
io
n
to
th
e
s
en
s
o
r
h
u
b
.
Af
ter
th
is
p
r
o
ce
s
s
,
th
e
d
ata
is
co
n
s
id
er
ed
clea
n
,
with
an
eq
u
al
n
u
m
b
er
o
f
en
tr
ies
f
o
r
ea
ch
co
lu
m
n
o
r
f
ea
t
u
r
e
v
ar
ia
b
le.
Min
-
Ma
x
s
ca
lin
g
is
p
er
f
o
r
m
ed
d
u
r
in
g
th
e
p
r
e
p
r
o
c
ess
in
g
p
h
ase
in
E
DA,
with
a
r
an
g
e
o
f
-
1
t
o
1
f
o
r
all
in
p
u
t
f
ea
tu
r
e
v
al
u
es.
T
h
e
p
u
r
p
o
s
e
o
f
th
e
Min
-
Ma
x
f
u
n
ct
io
n
is
to
en
s
u
r
e
d
ata
co
n
s
is
ten
cy
an
d
ac
ce
ler
ate
th
e
p
er
f
o
r
m
an
ce
o
f
alg
o
r
ith
m
s
d
u
r
i
n
g
t
h
e
l
e
a
r
n
i
n
g
p
r
o
c
e
s
s
.
T
h
i
s
n
o
r
m
a
l
i
z
a
ti
o
n
t
e
c
h
n
i
q
u
e
t
r
a
n
s
f
o
r
m
s
d
a
t
a
f
r
o
m
a
l
l
f
e
a
t
u
r
e
s
t
o
a
r
a
n
g
e
o
f
-
1
t
o
1
[
2
7
]
.
T
h
e
eq
u
atio
n
u
s
ed
in
th
i
s
r
esear
ch
is
as
f
o
llo
ws,
wh
er
e
is
th
e
h
ig
h
est
v
alu
e
an
d
is
th
e
lo
west
v
alu
e
in
th
e
s
am
p
le
d
ata.
=
−
−
×
(
1
−
(
−
1
)
)
+
(
−
1
)
(
4
)
Af
ter
n
o
r
m
alizin
g
all
d
ata
wit
h
in
th
e
r
an
g
e
o
f
-
1
to
1
,
th
e
n
e
x
t step
is
to
s
ep
ar
ate
th
e
in
p
u
t a
n
d
o
u
tp
u
t
d
ata
f
o
r
ea
c
h
s
m
ar
tp
lu
g
elec
tr
ical
ap
p
lian
ce
b
ased
o
n
th
e
co
r
r
elatio
n
c
o
ef
f
icien
t
(
1
)
.
I
n
th
is
s
tu
d
y
,
th
e
L
STM
tr
ain
in
g
p
r
o
ce
s
s
is
s
et
with
th
e
f
o
llo
win
g
p
ar
am
ete
r
s
:
2
5
ce
ll
o
r
n
e
u
r
o
n
s
in
ea
c
h
h
id
d
en
lay
e
r
(
s
ee
Fig
u
r
e
2
)
with
1
0
%
o
f
o
v
er
f
itti
n
g
,
7
5
%
o
f
th
e
d
ata
is
u
s
ed
f
o
r
tr
ain
in
g
,
2
5
%
f
o
r
test
in
g
,
2
5
ep
o
ch
s
,
a
lear
n
in
g
r
ate
o
f
0
.
0
0
0
1
,
a
n
d
th
e
o
p
tim
i
za
tio
n
alg
o
r
ith
m
u
s
ed
is
Ad
am
.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
is
r
esear
ch
in
v
esti
g
ates
th
e
im
p
ac
t
o
f
h
o
u
s
eh
o
l
d
b
eh
a
v
io
r
o
n
th
e
SHAS
to
p
r
ev
e
n
t
elec
tr
ical
o
v
er
lo
ad
.
W
h
ile
p
r
ev
io
u
s
s
tu
d
ies
also
aim
ed
to
p
r
e
v
en
t
o
v
er
lo
ad
,
t
h
eir
ap
p
r
o
ac
h
es
r
elied
o
n
d
eter
m
in
is
tic
f
ac
to
r
s
with
o
u
t
co
n
s
id
er
i
n
g
en
v
ir
o
n
m
en
tal
co
n
d
itio
n
s
th
a
t
in
f
lu
en
ce
h
o
u
s
eh
o
ld
b
eh
a
v
io
r
.
T
h
is
s
ec
tio
n
p
r
esen
ts
a
d
etailed
an
aly
s
is
o
f
th
e
r
esear
ch
f
in
d
in
g
s
,
s
tar
tin
g
with
ex
p
lo
r
at
o
r
y
d
ata
an
al
y
s
is
(
E
DA)
,
wh
ich
p
r
ep
r
o
ce
s
s
es
r
aw
d
ata
in
t
o
a
u
s
ab
le
d
ataset.
I
t
th
en
id
en
tifi
es
k
ey
i
n
p
u
t
v
ar
iab
les
f
o
r
th
e
L
STM
m
o
d
el
f
o
r
ea
ch
s
m
ar
t
p
lu
g
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
SHAS
is
ev
alu
ated
b
y
co
m
p
ar
i
n
g
r
esu
lts
f
r
o
m
d
if
f
er
en
t
d
ev
ices,
s
u
ch
as
S
-
AC
,
S
-
T
V,
an
d
S
-
W
PM,
u
s
i
n
g
r
ea
l
-
wo
r
ld
d
ata.
3
.
1
.
SH
AS
da
t
a
prepro
ce
s
s
ing
I
n
th
is
s
tu
d
y
,
d
ata
p
r
ep
r
o
ce
s
s
in
g
b
eg
in
s
with
ex
tr
ac
tin
g
th
e
AHE
AC
S
d
ataset
in
C
SV
f
o
r
m
at
f
r
o
m
th
e
d
ataset
r
ep
o
s
ito
r
y
.
T
h
e
d
at
a
is
th
en
tr
an
s
f
o
r
m
ed
an
d
n
o
r
m
alize
d
to
en
s
u
r
e
it
is
clea
n
,
with
n
o
n
u
ll
o
r
NaN
v
alu
es,
an
d
co
n
s
is
ten
t
in
s
ize,
m
ea
n
in
g
ea
c
h
f
ea
t
u
r
e
c
o
n
ta
in
s
th
e
s
am
e
n
u
m
b
e
r
o
f
d
ata
p
o
in
ts
.
A
f
ter
war
d
,
co
r
r
elatio
n
a
n
aly
s
is
is
p
er
f
o
r
m
ed
to
id
e
n
tify
t
h
e
in
p
u
t
f
e
atu
r
es
f
o
r
t
h
e
d
esire
d
s
m
ar
tp
lu
g
o
u
t
p
u
t
u
s
in
g
a
h
ea
tm
ap
v
is
u
aliza
tio
n
b
ased
o
n
(
1
)
.
Fig
u
r
e
3
s
h
o
ws
th
e
h
ea
t
m
ap
r
esu
lts
o
f
all
d
ataset
f
ea
t
u
r
es
o
r
v
ar
iab
les.
I
n
th
is
h
ea
tm
ap
,
a
v
alu
e
clo
s
e
to
1
(
d
ar
k
-
r
e
d
)
in
d
icate
s
a
s
tr
o
n
g
p
o
s
itiv
e
co
r
r
elatio
n
,
wh
ile
a
v
alu
e
clo
s
e
to
-
1
(
d
ar
k
b
lu
e)
in
d
icate
s
a
s
tr
o
n
g
n
eg
ativ
e
co
r
r
elatio
n
.
Fo
r
th
e
s
m
ar
tp
lu
g
co
n
n
ec
ted
to
th
e
AC
(
S
-
AC
)
,
1
5
ca
n
d
id
ate
in
p
u
t
f
ea
tu
r
es
wer
e
id
en
tifie
d
,
with
an
ab
s
o
lu
te
co
r
r
elatio
n
v
al
u
e
o
f
|
0
.
0
3
|
.
T
h
is
v
alu
e
was
ch
o
s
e
n
to
ac
c
o
u
n
t
f
o
r
th
e
tim
e
v
ar
i
ab
le,
in
f
lu
e
n
cin
g
th
e
S
-
AC
o
u
tp
u
t.
I
t
is
n
o
t
s
et
s
m
aller
to
av
o
id
i
n
v
o
lv
i
n
g
to
o
m
an
y
v
ar
ia
b
les
an
d
o
v
er
l
y
co
m
p
le
x
m
o
d
el.
B
ased
o
n
th
e
h
ea
tm
a
p
in
Fig
u
r
e
3
,
t
h
e
in
p
u
t
f
ea
tu
r
es
ar
e
[
T
P,
S
-
T
V,
S
-
W
PM,
S
-
MC,
T
,
H,
W
,
Pb
,
Pg
,
W
S]
wh
er
e
W
S
is
th
e
tim
e
v
ar
iab
le.
I
n
a
d
d
itio
n
to
S
-
AC
,
th
is
s
tu
d
y
also
ev
alu
ates
th
e
S
-
T
V
an
d
S
-
WP
M
s
m
ar
tp
lu
g
s
.
Fo
r
th
ese,
an
o
f
|
0
.
1
|
was
ch
o
o
s
en
.
T
h
e
in
p
u
t
f
ea
tu
r
es
f
o
r
S
-
T
V
ar
e
[
T
P,
S
-
AC
,
T
,
H,
W
,
Pg
,
NSM,
W
K]
,
an
d
f
o
r
S
-
W
PM,
th
e
in
p
u
t f
ea
t
u
r
es a
r
e
[
T
P,
S
-
A
C
,
S
-
MC,
T
,
W
]
.
W
e
f
o
u
n
d
th
at
tim
e
v
ar
iab
les
(
W
K,
HR
,
D,
W
S,
an
d
NSM)
h
av
e
an
in
f
lu
en
c
e
o
n
c
o
n
tr
o
l
v
ar
iab
les
s
u
ch
as
S
-
AC
,
S
-
T
V,
an
d
S
-
W
PM.
Alth
o
u
g
h
th
e
co
r
r
elati
o
n
s
ar
e
s
m
all,
r
an
g
in
g
f
r
o
m
0
.
0
1
to
0
.
1
7
,
tim
e
v
ar
iab
les
ex
tr
ac
ted
f
r
o
m
t
h
e
t
im
e
s
er
ies
d
ata
ar
e
p
ar
ticu
la
r
ly
s
ig
n
if
ican
t
d
u
e
to
t
h
eir
im
p
ac
t
o
n
h
o
u
s
eh
o
ld
ap
p
lian
ce
u
s
ag
e,
wh
ich
is
in
f
lu
en
ce
d
b
y
tim
e
-
b
ased
p
atter
n
s
.
T
h
e
r
esid
en
ts
'
s
p
ec
if
ic
r
o
u
tin
es
f
o
r
u
s
in
g
ap
p
lian
ce
s
lik
e
air
co
n
d
itio
n
er
s
o
r
wash
in
g
m
ac
h
in
es a
r
e
c
r
u
cial
f
o
r
ac
cu
r
ate
e
n
er
g
y
co
n
s
u
m
p
tio
n
p
r
ed
ictio
n
s
.
Alth
o
u
g
h
th
e
co
r
r
elatio
n
s
m
ay
ap
p
ea
r
lo
w,
ca
p
tu
r
i
n
g
th
ese
tem
p
o
r
al
p
atter
n
s
en
h
a
n
ce
s
an
aly
s
is
d
ep
th
co
m
p
ar
ed
to
s
tatic
d
eter
m
in
i
s
tic
co
n
tr
o
l
m
eth
o
d
s
[
1
9
]
,
[
2
0
]
.
T
h
e
f
lex
ib
ilit
y
in
d
ete
r
m
i
n
in
g
t
h
e
v
alu
e
allo
ws f
o
r
ad
ju
s
tm
en
ts
ac
co
r
d
in
g
to
th
e
d
esire
d
m
o
d
el
b
e
h
av
io
r
an
d
d
esig
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
a
r
y
20
25
:
7
5
8
-
7
7
0
764
Fig
u
r
e
3
.
C
o
ef
f
icie
n
t c
o
r
r
e
lati
o
n
b
etwe
en
f
ea
tu
r
e
in
AHE
AC
S d
ataset
Af
ter
d
eter
m
in
in
g
th
e
v
ar
iab
le
s
to
b
e
u
s
ed
as
in
p
u
t
f
ea
tu
r
es
f
o
r
ea
ch
s
m
ar
tp
lu
g
(
S
-
AC
,
S
-
T
V,
an
d
S
-
W
PM)
,
th
e
L
STM
m
o
d
el
lear
n
s
f
r
o
m
th
e
d
ataset
to
b
u
ild
a
class
if
icatio
n
m
o
d
el
with
o
u
tp
u
ts
o
f
eith
er
'
ON'
o
r
'
O
FF
'
f
o
r
ea
ch
s
m
ar
tp
lu
g
.
T
a
b
le
2
s
h
o
ws
th
e
r
esu
lts
o
f
th
e
lear
n
in
g
p
r
o
ce
s
s
co
n
d
u
cted
i
n
th
e
G
o
o
g
le
C
o
lab
en
v
ir
o
n
m
en
t.
T
h
e
co
r
r
elatio
n
b
etwe
en
Fig
u
r
e
2
,
wh
ic
h
d
ep
i
cts
th
e
L
STM
d
esig
n
,
an
d
T
ab
le
2
,
wh
ich
p
r
esen
ts
th
e
r
esu
lts
o
f
th
e
im
p
lem
en
tatio
n
u
s
in
g
Py
th
o
n
,
s
h
o
ws
th
at
th
is
m
o
d
el
h
as
th
r
ee
lay
er
s
:
L
STM
,
Dr
o
p
o
u
t,
a
n
d
Den
s
e
(
Ou
tp
u
t)
.
Fo
r
ex
am
p
le,
th
e
im
p
lem
en
tatio
n
f
o
r
S
-
AC
i
s
ex
p
lain
ed
as f
o
llo
ws:
a)
ls
tm
_
1
:
T
h
is
is
th
e
f
ir
s
t
L
STM
lay
er
in
th
e
m
o
d
el,
with
a
n
o
u
tp
u
t
s
h
ap
e
o
f
(
No
n
e,
1
0
,
2
5
)
.
T
h
is
m
ea
n
s
th
at
th
e
lay
er
h
as
2
5
u
n
its
(
o
r
m
em
o
r
y
ce
lls
)
,
an
d
it
p
r
o
d
u
c
es
a
s
eq
u
en
ce
o
f
1
0
o
u
tp
u
t
v
ec
to
r
s
,
wh
er
e
ea
ch
v
ec
to
r
h
as
2
5
elem
en
ts
.
T
h
ese
o
u
tp
u
t
v
ec
to
r
s
ar
e
eq
u
i
v
alen
t
to
th
e
n
u
m
b
er
o
f
in
p
u
t
f
ea
tu
r
es:
S
-
AC
h
as
1
0
in
p
u
t
f
ea
tu
r
es,
S
-
T
V
h
as
8
in
p
u
t
f
ea
tu
r
es,
wh
ile
S
-
WP
M
h
as
5
in
p
u
t
f
ea
tu
r
es.
T
h
e
L
STM
lay
er
r
ec
eiv
es in
p
u
t f
r
o
m
th
e
in
p
u
t l
ay
er
an
d
p
r
o
ce
s
s
es it b
y
u
p
d
at
in
g
its
in
ter
n
al
s
tate,
wh
ich
co
n
s
is
ts
o
f
a
ce
ll
s
tate
an
d
a
h
id
d
en
s
tate.
T
h
e
L
STM
lay
er
ca
n
lear
n
to
k
ee
p
o
r
d
is
ca
r
d
in
f
o
r
m
atio
n
f
r
o
m
th
e
p
r
ev
io
u
s
tim
e
s
t
ep
s
,
d
ep
en
d
in
g
o
n
its
r
elev
an
ce
f
o
r
p
r
e
d
ictin
g
th
e
o
u
tp
u
t.
b)
d
r
o
p
o
u
t_
1
: T
h
is
is
a
d
r
o
p
o
u
t l
ay
er
,
wh
ich
is
u
s
ed
to
p
r
ev
en
t o
v
er
f
itti
n
g
b
y
r
an
d
o
m
ly
s
ettin
g
a
f
r
ac
tio
n
o
f
th
e
in
p
u
t
u
n
its
to
ze
r
o
d
u
r
in
g
tr
ain
in
g
.
T
h
e
d
r
o
p
o
u
t
r
ate
is
d
en
o
ted
b
y
1
0
%
o
r
0
.
1
,
as
m
en
tio
n
ed
b
e
f
o
r
e
in
th
e
p
r
ev
io
u
s
s
ec
tio
n
.
T
h
e
o
u
tp
u
t
s
h
ap
e
o
f
th
e
d
r
o
p
o
u
t
la
y
er
is
th
e
s
am
e
as
th
e
in
p
u
t
s
h
ap
e,
wh
ic
h
is
(
No
n
e,
1
0
,
2
5
)
in
t
h
is
ca
s
e.
c)
ls
tm
_
2
:
T
h
is
is
th
e
s
ec
o
n
d
L
S
T
M
lay
er
in
th
e
m
o
d
el,
with
an
o
u
tp
u
t
s
h
ap
e
o
f
(
No
n
e,
2
5
)
.
T
h
is
m
ea
n
s
th
at
th
e
l
ay
er
h
as
2
5
u
n
its
,
an
d
it
p
r
o
d
u
ce
s
a
s
in
g
le
o
u
tp
u
t
v
ec
to
r
f
o
r
ea
c
h
in
p
u
t
s
eq
u
e
n
c
e.
T
h
is
L
STM
lay
er
r
ec
eiv
es
in
p
u
t
f
r
o
m
th
e
p
r
ev
io
u
s
lay
er
an
d
p
r
o
ce
s
s
es
it
b
y
u
p
d
atin
g
its
in
ter
n
al
s
tate,
wh
ich
co
n
s
is
ts
o
f
a
ce
ll st
ate
an
d
a
h
i
d
d
en
s
tate.
d)
d
r
o
p
o
u
t_
2
:
T
h
is
is
an
o
th
er
d
r
o
p
o
u
t
lay
er
,
wh
ich
is
u
s
ed
to
p
r
ev
en
t
o
v
er
f
itti
n
g
b
y
r
an
d
o
m
ly
s
ettin
g
a
f
r
ac
tio
n
o
f
th
e
in
p
u
t
u
n
its
to
ze
r
o
d
u
r
in
g
tr
ai
n
in
g
.
T
h
e
d
r
o
p
o
u
t
r
ate
is
d
en
o
ted
b
y
1
0
%
an
d
th
e
o
u
tp
u
t
s
h
ap
e
o
f
th
is
d
r
o
p
o
u
t la
y
er
is
t
h
e
s
am
e
as th
e
in
p
u
t sh
a
p
e,
w
h
ich
is
(
No
n
e,
2
5
)
i
n
th
is
ca
s
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
N
a
tu
r
a
l sma
r
t h
o
me
a
u
to
ma
tio
n
s
ystem
u
s
in
g
LS
TM
b
a
s
ed
o
n
h
o
u
s
eh
o
ld
b
eh
a
vi
o
u
r
(
Mo
c
h
a
ma
d
S
u
s
a
n
to
k
)
765
e)
d
en
s
e:
T
h
is
is
th
e
o
u
tp
u
t
lay
e
r
o
f
th
e
m
o
d
el,
wh
ich
is
a
f
u
l
ly
co
n
n
ec
ted
la
y
er
with
1
u
n
i
t
(
o
r
n
eu
r
o
n
)
.
T
h
e
o
u
tp
u
t
s
h
ap
e
o
f
th
e
d
en
s
e
lay
er
is
(
No
n
e,
1
)
,
wh
ic
h
m
e
an
s
th
at
it
p
r
o
d
u
ce
s
a
s
in
g
le
s
ca
lar
v
alu
e
f
o
r
ea
ch
in
p
u
t
s
eq
u
en
ce
.
T
h
e
ac
tiv
atio
n
f
u
n
ctio
n
o
f
th
e
d
en
s
e
lay
er
is
tan
h
,
wh
ich
is
u
s
ed
f
o
r
b
in
ar
y
class
if
icatio
n
task
s
.
T
ab
le
2
.
Su
m
m
a
r
y
L
STM
clas
s
if
icatio
n
m
o
d
el
La
y
e
r
S
-
AC
S
-
TV
S
-
WPM
O
u
t
p
u
t
S
h
a
p
e
P
a
r
a
m
#
O
u
t
p
u
t
S
h
a
p
e
P
a
r
a
m
#
O
u
t
p
u
t
S
h
a
p
e
P
a
r
a
m
#
l
st
m
_
1
(
N
o
n
e
,
1
0
,
2
5
)
2
7
0
0
(
N
o
n
e
,
8
,
2
5
)
2
7
0
0
(
N
o
n
e
,
5
,
2
5
)
2
7
0
0
d
r
o
p
o
u
t
_
1
(
N
o
n
e
,
1
0
,
2
5
)
0
(
N
o
n
e
,
8
,
2
5
)
0
(
N
o
n
e
,
5
,
2
5
)
0
l
st
m
_
2
(
N
o
n
e
,
2
5
)
5
1
0
0
(
N
o
n
e
,
2
5
)
5
1
0
0
(
N
o
n
e
,
2
5
)
5
1
0
0
d
r
o
p
o
u
t
_
2
(
N
o
n
e
,
2
5
)
0
(
N
o
n
e
,
2
5
)
0
(
N
o
n
e
,
2
5
)
0
d
e
n
se
(
N
o
n
e
,
1
)
26
(
N
o
n
e
,
1
)
26
(
N
o
n
e
,
1
)
26
To
t
a
l
p
a
r
a
ms
:
7
8
2
6
Tr
a
i
n
a
b
l
e
p
a
r
a
ms
:
7
8
0
0
N
o
n
-
t
r
a
i
n
a
b
l
e
p
a
r
a
ms:
2
6
T
h
e
to
tal
n
u
m
b
er
o
f
p
ar
am
eter
s
in
th
e
m
o
d
el
is
7
,
8
2
6
,
wh
ich
in
clu
d
es 7
,
8
0
0
tr
ain
ab
le
p
ar
a
m
eter
s
an
d
2
6
n
o
n
-
tr
ain
ab
le
p
ar
am
eter
s
.
T
h
e
tr
ain
a
b
le
p
a
r
am
eter
s
ar
e
th
e
weig
h
ts
an
d
b
iases
o
f
th
e
L
STM
an
d
d
en
s
e
lay
er
s
,
wh
ich
ar
e
lear
n
e
d
d
u
r
in
g
tr
ain
in
g
.
T
h
e
n
o
n
-
tr
ain
a
b
l
e
p
ar
am
eter
s
ar
e
th
e
d
r
o
p
o
u
t
m
ask
s
,
wh
ich
ar
e
r
an
d
o
m
l
y
g
e
n
er
ated
an
d
f
ix
ed
d
u
r
in
g
tr
ai
n
in
g
.
T
h
ese
r
esu
lt
s
d
em
o
n
s
tr
ate
th
e
r
eliab
ilit
y
o
f
L
STM
as
o
n
e
o
f
th
e
b
est
-
p
er
f
o
r
m
in
g
d
ee
p
lear
n
in
g
alg
o
r
ith
m
s
,
p
ar
ticu
lar
ly
i
n
h
an
d
lin
g
lar
g
e
-
s
ca
le
d
ata
s
u
ch
as
in
th
is
s
tu
d
y
wh
ich
u
s
es
a
m
o
d
el
with
5
l
ay
er
s
(
1
in
p
u
t
lay
er
,
3
h
id
d
e
n
lay
er
s
,
an
d
1
o
u
t
p
u
t
lay
er
)
[
2
8
]
.
H
o
wev
er
,
th
e
L
STM
d
esig
n
in
th
is
s
tu
d
y
,
as
p
r
ev
io
u
s
ly
ex
p
lain
ed
,
m
ai
n
t
ain
s
s
im
p
licity
to
en
s
u
r
e
it
s
a
p
p
licab
ilit
y
o
n
ed
g
e
d
ev
ices
with
in
I
o
T
n
etwo
r
k
s
.
Fu
r
th
er
r
e
f
in
e
L
STM
ar
ch
itec
tu
r
es
to
r
ed
u
ce
co
m
p
u
tatio
n
al
co
m
p
le
x
ity
wh
ile
m
ain
tain
in
g
o
r
ev
e
n
im
p
r
o
v
i
n
g
ac
cu
r
ac
y
.
R
esear
ch
co
u
ld
f
o
cu
s
o
n
p
r
u
n
in
g
tech
n
i
q
u
es,
q
u
an
tizatio
n
,
an
d
ef
f
icien
t
m
em
o
r
y
m
an
a
g
em
e
n
t
to
m
ak
e
L
STM
m
o
d
els
e
v
en
m
o
r
e
s
u
itab
le
f
o
r
e
d
g
e
d
ev
ices
with
lim
ited
r
eso
u
r
ce
s
.
3
.
2
.
SH
AS
perf
o
r
m
a
nce
a
na
ly
s
is
T
h
is
s
tu
d
y
u
s
es
L
STM
in
a
SHAS
ap
p
licatio
n
to
p
r
o
d
u
ce
o
u
tp
u
t
in
th
e
f
o
r
m
o
f
O
N
o
r
OF
F
class
if
icatio
n
s
.
T
h
ese
ON
o
r
OFF
m
ess
ag
es
ar
e
th
en
f
o
r
w
ar
d
ed
t
o
th
e
s
m
ar
tp
lu
g
v
ia
th
e
MQ
T
T
p
r
o
to
co
l,
wh
er
e
th
ey
ar
e
tr
a
n
s
lated
b
y
t
h
e
r
elay
in
th
e
s
m
ar
tp
l
u
g
in
to
an
ON
o
r
OFF s
witch
to
co
n
tr
o
l th
e
p
o
w
er
f
l
o
w
to
th
e
elec
tr
ical
ap
p
lian
ce
.
T
o
ev
alu
ate
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
L
STM
m
o
d
el
in
th
is
SHAS,
th
e
co
n
f
u
s
io
n
m
atr
ix
m
eth
o
d
is
u
s
ed
,
wh
ich
is
a
s
tan
d
ar
d
ap
p
r
o
ac
h
in
class
if
icatio
n
m
o
d
elin
g
.
T
h
e
c
o
n
f
u
s
io
n
m
atr
ix
=
(
)
p
r
o
v
id
es
i
n
s
ig
h
t
in
to
t
h
e
m
o
d
el'
s
ac
cu
r
ac
y
[
2
9
]
.
Her
e,
T
r
u
e
Po
s
itiv
e
(
TP
)
r
ep
r
esen
t
s
ca
s
es
wh
er
e
th
e
s
m
ar
tp
lu
g
is
co
r
r
ec
tly
p
r
ed
ict
ed
to
b
e
ON,
Fals
e
Neg
ativ
e
(
FN
)
r
ep
r
esen
ts
ca
s
es
wh
er
e
th
e
p
lu
g
is
p
r
ed
icted
to
b
e
OFF
wh
en
it
is
ac
tu
ally
ON,
T
r
u
e
Neg
ativ
e
(
TN
)
r
ep
r
esen
ts
co
r
r
ec
t
p
r
ed
ictio
n
s
th
at
th
e
p
lu
g
is
O
FF
,
an
d
Fals
e
Po
s
itiv
e
(
FP
)
r
ep
r
esen
ts
in
co
r
r
ec
t
p
r
ed
ictio
n
s
wh
e
r
e
th
e
p
lu
g
is
p
r
e
d
icted
to
b
e
ON
wh
en
it
is
OFF.
All
th
e
p
r
ed
icted
d
ata
f
r
o
m
th
e
L
STM
m
o
d
el,
o
r
d
er
ed
b
y
tim
e,
will
b
e
cla
s
s
if
ied
in
to
t
h
ese
f
o
u
r
v
ar
iab
les.
B
elo
w
ar
e
th
e
r
esu
lts
o
f
th
e
c
o
n
f
u
s
io
n
m
atr
ix
f
r
o
m
th
e
L
STM
m
o
d
el
with
a
to
tal
o
f
8
0
,
8
1
8
d
ata
p
o
in
ts
o
n
th
e
d
ataset
f
o
r
ea
ch
S
-
AC
,
S
-
T
V,
an
d
S
-
W
PM,
wh
ich
wer
e
r
u
n
i
n
th
e
Go
o
g
le
C
o
lab
en
v
i
r
o
n
m
en
t u
s
in
g
Py
th
o
n
:
−
=
(
40594
4339
10418
25467
)
−
=
(
22273
450
1155
56940
)
−
=
(
5728
1044
952
73094
)
Nex
t,
th
e
co
n
f
u
s
io
n
m
atr
i
x
p
ar
am
eter
s
ar
e
ca
lcu
lated
u
s
in
g
t
h
e
(
5
)
-
(
8
)
[
3
0
]
:
=
+
+
+
+
(
5
)
=
+
(
6
)
=
+
(
7
)
1
−
=
2
∗
∗
+
(
8
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
7
,
No
.
2
,
Feb
r
u
a
r
y
20
25
:
7
5
8
-
7
7
0
766
T
h
e
p
e
r
f
o
r
m
an
c
e
e
v
alu
atio
n
o
f
th
e
SHAS
u
s
in
g
(
5
)
-
(
8
)
i
s
s
u
m
m
ar
ized
i
n
T
ab
le
3
an
d
v
is
u
ally
r
ep
r
esen
ted
in
Fig
u
r
e
4
.
T
h
e
e
v
alu
atio
n
in
d
icate
s
th
at
th
e
L
STM
m
o
d
el'
s
lear
n
in
g
o
f
th
e
h
o
u
s
eh
o
ld
b
eh
av
io
r
is
q
u
ite
s
u
cc
es
s
f
u
l,
ac
h
iev
in
g
h
ig
h
ac
cu
r
ac
y
f
o
r
th
e
s
m
ar
tp
l
u
g
T
V
(
S
-
T
V)
at
9
8
%
an
d
th
e
s
m
ar
tp
lu
g
W
ater
Pu
m
p
Ma
ch
in
e
(
S
-
W
PM)
at
9
7
.
6
%,
b
u
t
s
lig
h
tly
lo
wer
f
o
r
th
e
s
m
ar
tp
lu
g
AC
(
S
-
AC
)
at
8
1
.
9
%.
Fo
r
S
-
T
V,
p
r
ec
is
io
n
an
d
r
ec
all
ar
e
b
o
t
h
v
er
y
h
i
g
h
at
9
5
%
an
d
9
8
%,
r
esp
ec
tiv
ely
.
T
h
is
in
d
icate
s
th
at
th
e
L
STM
m
o
d
el
ac
cu
r
ately
p
r
ed
icts
th
e
ON/O
FF
s
tatu
s
o
f
th
e
T
V,
with
m
in
i
m
al
f
alse
p
o
s
itiv
es
(
FP
)
an
d
f
alse
n
eg
ativ
es
(
FN)
.
S
-
W
PM
also
d
em
o
n
s
tr
ates
g
o
o
d
p
er
f
o
r
m
a
n
ce
,
with
a
p
r
ec
is
io
n
o
f
8
5
.
7
%
an
d
r
ec
all
o
f
8
5
%,
m
ea
n
in
g
th
at
m
o
s
t
in
s
tan
ce
s
ar
e
co
r
r
ec
tly
d
etec
ted
,
th
o
u
g
h
a
f
ew
ar
e
m
is
s
ed
.
I
n
co
n
tr
ast,
S
-
AC
s
h
o
ws
a
lo
wer
p
r
ec
is
io
n
o
f
7
9
.
5
%
c
o
m
p
ar
e
d
to
a
r
ec
all
o
f
9
0
.
3
%,
s
u
g
g
esti
n
g
th
at
wh
ile
th
e
m
o
d
el
c
o
r
r
ec
tly
p
r
ed
icts
m
o
s
t
S
-
AC
in
s
tan
ce
s
,
th
er
e
ar
e
m
o
r
e
f
als
e
p
o
s
itiv
es.
T
h
e
p
er
f
o
r
m
an
ce
g
ap
b
etwe
en
S
-
T
V
an
d
S
-
A
C
co
u
ld
b
e
d
u
e
to
v
ar
io
u
s
f
ac
to
r
s
af
f
ec
tin
g
ap
p
lian
ce
u
s
ag
e
p
atter
n
s
.
T
h
e
ac
cu
r
ate
p
r
ed
ictio
n
o
f
s
m
ar
tp
lu
g
o
p
e
r
atio
n
s
,
p
ar
ticu
lar
ly
f
o
r
S
-
T
V
an
d
S
-
W
PM,
d
ir
ec
tly
co
n
tr
ib
u
tes
to
p
r
ev
en
tin
g
elec
t
r
ical
o
v
er
lo
a
d
s
.
T
h
is
p
r
ev
en
tio
n
is
cr
u
cial
f
o
r
m
ain
tain
in
g
h
o
u
s
eh
o
ld
s
af
ety
b
y
r
e
d
u
cin
g
th
e
r
is
k
o
f
MCB
tr
ip
s
,
wh
ich
ca
n
lead
to
eq
u
ip
m
en
t
d
am
ag
e
o
r
s
af
ety
h
az
a
r
d
s
.
Ad
d
itio
n
ally
,
th
e
s
y
s
tem
en
s
u
r
es
en
er
g
y
m
a
n
ag
em
e
n
t
is
o
p
tim
ized
b
y
o
n
ly
ac
tiv
atin
g
ap
p
lian
ce
s
wh
e
n
n
e
ce
s
s
ar
y
,
b
ased
o
n
lear
n
ed
u
s
er
h
ab
its
.
Fu
tu
r
e
r
esear
ch
co
u
l
d
ex
p
lo
r
e
co
n
tex
tu
al
an
d
e
n
v
ir
o
n
m
en
tal
f
ac
to
r
s
,
s
u
ch
as
wea
th
er
co
n
d
itio
n
s
,
tim
e
o
f
d
ay
,
a
n
d
u
s
er
b
eh
av
io
r
p
atter
n
s
,
th
at
im
p
ac
t
a
p
p
lian
ce
u
s
ag
e.
Ad
d
itio
n
ally
,
th
e
ac
cu
r
ac
y
o
f
s
en
s
o
r
s
,
l
ik
e
th
o
s
e
d
etec
tin
g
r
o
o
m
o
cc
u
p
an
cy
,
co
u
ld
b
e
an
o
th
e
r
k
ey
f
ac
to
r
co
n
tr
i
b
u
tin
g
to
t
h
is
g
ap
.
I
m
p
r
o
v
in
g
s
en
s
o
r
ac
cu
r
ac
y
o
r
u
s
in
g
s
en
s
o
r
f
u
s
io
n
tech
n
i
q
u
es
co
u
l
d
en
h
an
ce
m
o
d
el
p
r
ed
ictio
n
s
,
esp
ec
ially
f
o
r
ap
p
lian
ce
s
lik
e
AC
th
at
ar
e
s
en
s
itiv
e
to
r
o
o
m
o
cc
u
p
an
c
y
.
T
ab
le
3
.
T
h
e
SHAS
p
er
f
o
r
m
a
n
ce
P
e
r
f
o
r
ma
n
c
e
M
e
t
r
i
c
s
S
martp
l
u
g
S
-
AC
S
-
TV
S
-
WPM
A
c
c
u
r
a
c
y
(
%)
8
1
.
9
98
9
7
.
6
P
r
e
c
i
ss
i
o
n
(
%)
7
9
.
5
95
8
5
.
7
R
e
c
a
l
l
(
%)
9
0
.
3
98
85
F
1
S
c
o
r
e
(
%)
85
97
85
Fig
u
r
e
4
.
SHAS C
o
n
f
u
s
io
n
m
atr
ix
p
er
f
o
r
m
an
ce
with
co
m
p
a
r
is
o
n
o
f
th
r
ee
s
m
ar
tp
lu
g
s
3
.
3
.
SH
AS
perf
o
r
m
a
nce
ev
a
lua
t
io
n wit
h r
ea
l
-
wo
rld da
t
a
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
a
b
o
v
e
m
o
d
el
was
f
u
r
t
h
er
test
ed
with
o
u
t
-
of
-
s
am
p
le
d
ata
o
r
ex
ter
n
al
v
alid
atio
n
d
ata,
s
p
ec
if
ically
d
ata
o
u
ts
id
e
th
e
r
an
g
e
Feb
r
u
ar
y
to
Ap
r
il
2
0
2
4
as
d
escr
ib
e
d
in
th
e
m
eth
o
d
o
l
o
g
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
N
a
tu
r
a
l sma
r
t h
o
me
a
u
to
ma
tio
n
s
ystem
u
s
in
g
LS
TM
b
a
s
ed
o
n
h
o
u
s
eh
o
ld
b
eh
a
vi
o
u
r
(
Mo
c
h
a
ma
d
S
u
s
a
n
to
k
)
767
s
ec
tio
n
.
T
h
is
s
tu
d
y
u
s
ed
ex
ter
n
al
v
alid
atio
n
d
ata
f
r
o
m
th
e
p
er
io
d
o
f
Ma
y
1
4
-
2
1
,
2
0
2
4
,
with
a
to
tal
o
f
5
,
5
2
0
d
ata
p
o
i
n
ts
.
Af
ter
p
r
e
-
p
r
o
ce
s
s
in
g
,
5
,
1
0
9
d
ata
p
o
i
n
ts
r
em
ain
ed
,
as
s
h
o
wn
in
Fig
u
r
e
5
.
Fig
u
r
e
5
(
a)
s
h
o
ws
th
e
f
ir
s
t
6
r
o
ws
o
f
th
e
d
ataset,
in
d
icatin
g
s
o
m
e
NaN
v
alu
es.
F
o
llo
win
g
p
r
e
-
p
r
o
ce
s
s
in
g
,
in
cl
u
d
in
g
tr
an
s
f
o
r
m
atio
n
o
f
th
e
T
S
c
o
lu
m
n
an
d
n
o
r
m
ali
za
tio
n
,
th
e
d
ata
b
ec
am
e
co
n
s
is
ten
t
ac
r
o
s
s
all
co
lu
m
n
s
as d
escr
ib
e
in
Fig
u
r
e
5
(
b
)
d
an
5
(
c
)
.
(
a)
(
b
)
(
c)
Fig
u
r
e
5
.
Pre
-
p
r
o
ce
s
s
in
g
ex
te
r
n
al
v
alid
atio
n
d
ata
:
(
a)
in
itial
o
f
ex
ter
n
al
v
alid
atio
n
d
ata
,
(
b
)
ex
ter
n
al
v
alid
atio
n
d
ata
with
ex
tr
ac
ted
T
S
,
a
n
d
(
c
)
ex
ter
n
al
v
alid
atio
n
d
ata
a
f
ter
n
o
r
m
lizatio
n
a
n
d
clea
n
in
g
T
h
e
L
STM
m
o
d
el
was
t
h
en
te
s
ted
o
n
t
h
is
ex
ter
n
al
v
alid
atio
n
d
ata
f
o
r
ea
ch
s
m
ar
tp
lu
g
(
S
-
AC
,
S
-
T
V,
an
d
S
-
W
PM)
,
with
ac
c
u
r
ac
y
r
esu
lts
s
h
o
wn
in
T
ab
le
4
a
n
d
Fig
u
r
e
6
.
T
h
e
test
was
d
iv
id
ed
in
to
t
h
r
ee
ca
teg
o
r
ies:
wee
k
d
ay
s
(
Mo
n
d
ay
to
Fri
d
ay
)
,
wee
k
en
d
s
(
Satu
r
d
ay
an
d
Su
n
d
a
y
)
,
an
d
a
ll
d
ay
s
co
m
b
in
ed
.
Acc
u
r
ac
y
f
o
r
S
-
AC
was
lo
wer
co
m
p
ar
ed
to
S
-
T
V
an
d
S
-
W
PM,
p
ar
ticu
lar
ly
d
u
r
in
g
wee
k
d
ay
s
,
wh
er
e
S
-
AC
ac
h
iev
ed
5
1
%,
S
-
T
V
6
6
.
1
%,
an
d
S
-
W
PM
8
3
.
3
%.
I
n
ter
esti
n
g
ly
,
th
e
ac
cu
r
ac
y
o
f
s
m
ar
tp
l
u
g
S
-
AC
in
cr
ea
s
es
d
u
r
in
g
wee
k
e
n
d
s
co
m
p
a
r
ed
t
o
wee
k
d
ay
s
,
r
ea
ch
in
g
6
7
.
3
%,
wh
ile
f
o
r
S
-
W
PM,
it
r
em
ain
s
r
elativ
ely
s
tab
le
at
ar
o
u
n
d
8
2
.
2
%.
A
n
o
th
er
in
ter
e
s
tin
g
o
b
s
er
v
atio
n
is
th
at
S
-
T
V
ac
h
iev
in
g
its
lo
west
ac
cu
r
ac
y
o
f
o
n
ly
4
7
.
5
%
o
n
wee
k
en
d
s
b
u
t
f
air
in
wee
k
d
a
y
s
at
6
6
.
1
%.
W
h
en
c
o
n
s
id
er
i
n
g
all
d
ay
s
,
S
-
W
PM
co
n
s
is
ten
tly
s
h
o
ws
h
ig
h
er
ac
cu
r
ac
y
c
o
m
p
ar
e
d
to
S
-
AC
an
d
S
-
T
V
at
a
n
av
e
r
ag
e
s
co
r
e
8
2
.
8
%.
T
h
is
s
u
g
g
ests
th
at
th
e
m
o
d
el
m
a
y
b
e
b
etter
at
p
r
ed
ictin
g
th
e
b
eh
av
io
r
o
f
th
e
wate
r
p
u
m
p
m
ac
h
in
e
co
m
p
a
r
ed
t
o
th
e
air
co
n
d
itio
n
er
an
d
telev
is
io
n
.
Ad
d
itio
n
ally
,
th
e
m
o
d
elin
g
o
f
S
-
W
PM
is
n
o
t
i
n
f
lu
en
ce
d
b
y
tim
e
f
ac
to
r
s
,
as
ev
id
e
n
ce
d
b
y
th
e
lo
w
co
r
r
elatio
n
co
ef
f
icien
t
v
alu
es
f
o
r
S
-
W
PM
with
in
p
u
t
f
ea
tu
r
e
v
ar
iab
les
in
th
e
tim
e
ca
teg
o
r
y
s
u
ch
as
[
W
K,
HR
,
D,
W
S,
NSM]
,
with
a
v
alu
e
o
f
|
0
.
0
1
|
(
r
ef
er
b
ac
k
to
Fig
u
r
e
3
)
.
T
h
ese
v
ar
iatio
n
s
in
ac
cu
r
ac
y
,
p
ar
ticu
lar
ly
f
o
r
S
-
AC
an
d
S
-
T
V,
h
ig
h
lig
h
t
th
e
im
p
o
r
tan
ce
o
f
c
o
n
s
id
er
in
g
tim
e
-
b
as
ed
u
s
er
h
ab
its
in
p
r
ev
en
tin
g
o
v
er
lo
a
d
s
.
Du
r
in
g
wee
k
en
d
s
,
wh
en
u
s
ag
e
p
atter
n
s
ar
e
less
p
r
ed
ictab
le,
th
e
s
y
s
tem
's
ab
ilit
y
to
ad
ju
s
t
to
th
es
e
ch
an
g
es
b
ec
o
m
es
cr
itical
f
o
r
b
o
th
s
af
ety
an
d
e
n
er
g
y
e
f
f
icien
cy
.
B
y
d
y
n
am
i
ca
lly
ad
ap
tin
g
t
o
d
if
f
er
en
t
s
c
en
ar
io
s
,
th
e
s
y
s
tem
en
s
u
r
es
th
at
ap
p
lian
ce
s
ar
e
n
o
t
lef
t
r
u
n
n
in
g
u
n
n
ec
ess
ar
ily
,
th
er
eb
y
r
ed
u
cin
g
t
h
e
r
is
k
o
f
o
v
e
r
lo
ad
s
an
d
o
p
tim
izin
g
p
o
wer
co
n
s
u
m
p
tio
n
.
T
ab
le
4
.
SHAS
p
er
f
o
r
m
an
ce
u
s
in
g
ex
ter
n
al
v
alid
atio
n
d
ata
P
e
r
f
o
r
ma
n
c
e
me
t
r
i
c
s
Ty
p
e
o
f
d
a
t
a
S
martp
l
u
g
S
-
AC
S
-
TV
S
-
WPM
A
c
c
u
r
a
c
y
(
%)
W
e
e
k
d
a
y
s
51
6
6
.
1
8
3
.
3
W
e
e
k
e
n
d
6
7
.
3
4
7
,
5
8
2
.
2
A
l
l
D
a
y
s
55
6
0
.
8
83
A
v
e
r
a
g
e
o
f
a
c
c
u
r
a
c
y
(
%)
5
7
.
8
5
8
.
1
8
2
.
8
Evaluation Warning : The document was created with Spire.PDF for Python.