I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
,
p
p
.
270
~
285
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
14
.i
1
.
p
p
2
7
0
-
2
8
5
270
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
M
a
chine learning
-
a
ss
isted decisio
n
supp
o
rt
in
indust
ria
l
ma
nufacturing
:
a ca
se study
on i
nj
ection mo
lding
m
a
chine
selection
F
a
o
uzi Ta
y
a
la
t
i,
So
ula
i
m
a
ne
I
diri,
I
k
hla
s
s
B
o
uk
ro
uh
,
Ab
d
el
la
h Az
m
a
ni,
M
o
nir A
zm
a
ni
I
n
t
e
l
l
i
g
e
n
t
A
u
t
o
m
a
t
i
o
n
a
n
d
B
i
o
M
e
d
G
e
n
o
mi
c
s
L
a
b
o
r
a
t
o
r
y
,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
o
f
T
a
n
g
i
e
r
,
A
b
d
e
l
m
a
l
e
k
Essa
a
d
i
U
n
i
v
e
r
s
i
t
y
,
Te
t
o
u
a
n
,
M
o
r
o
c
c
o
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Mar
21
,
2
0
2
4
R
ev
is
ed
J
u
n
23
,
2
0
2
4
Acc
ep
ted
J
u
l
26
,
2
0
2
4
S
e
lec
ti
n
g
th
e
ri
g
h
t
i
n
jec
ti
o
n
m
o
l
d
in
g
m
a
c
h
in
e
fo
r
n
e
w
p
ro
d
u
c
ts
re
m
a
in
s
a
c
h
a
ll
e
n
g
i
n
g
tas
k
th
a
t
sig
n
ifi
c
a
n
tl
y
i
n
flu
e
n
c
e
s
th
e
p
r
o
fit
a
b
il
it
y
a
n
d
flex
i
b
il
it
y
o
f
c
o
m
p
a
n
ies
.
Th
e
c
o
n
v
e
n
ti
o
n
a
l
a
p
p
r
o
a
c
h
in
v
o
l
v
e
s
p
e
rfo
rm
in
g
t
h
e
o
re
ti
c
a
l
c
a
lcu
latio
n
s
f
o
r
c
lam
p
i
n
g
fo
rc
e
,
c
o
n
d
u
c
ti
n
g
m
e
c
h
a
n
ica
l
v
a
li
d
a
ti
o
n
s
o
f
th
e
m
o
ld
,
a
n
d
c
a
rry
in
g
o
u
t
re
a
l
tri
a
ls
fo
r
n
e
w
p
a
rts.
T
h
is
a
p
p
r
o
a
c
h
is
ti
m
e
-
c
o
n
su
m
in
g
,
c
o
stly
,
a
n
d
re
q
u
ires
a
h
ig
h
lev
e
l
o
f
e
x
p
e
rti
se
to
e
n
su
re
t
h
e
o
p
ti
m
a
l
m
a
c
h
in
e
c
h
o
ice
.
Th
is
st
u
d
y
e
x
p
lo
re
s
th
e
u
se
o
f
m
a
c
h
in
e
lea
rn
in
g
(M
L)
m
e
th
o
d
s
fo
r
e
fficie
n
t
m
a
c
h
in
e
se
lec
ti
o
n
b
a
se
d
o
n
p
ro
d
u
c
t,
m
a
teria
l,
a
n
d
m
o
ld
c
rit
e
ria.
S
ix
su
p
e
rv
ise
d
lea
rn
i
n
g
tec
h
n
i
q
u
e
s
we
re
tes
ted
o
n
a
d
a
tas
e
t
c
o
m
p
risin
g
7
0
p
las
ti
c
p
a
rts
a
n
d
f
iv
e
m
a
c
h
in
e
s.
Ev
a
lu
a
ti
o
n
m
e
tri
c
s
li
k
e
F1
-
sc
o
re
,
re
c
a
ll
,
p
re
c
isio
n
,
a
n
d
a
c
c
u
ra
c
y
we
re
u
se
d
to
c
o
m
p
a
re
m
o
d
e
ls.
Th
e
re
su
lt
s
i
n
d
ica
te
t
h
a
t
ML
c
a
n
p
ro
v
id
e
g
u
id
a
n
c
e
fo
r
p
re
d
ictin
g
m
a
c
h
in
e
se
lec
ti
o
n
,
with
a
p
re
fe
re
n
c
e
fo
r
t
h
e
ra
n
d
o
m
fo
re
st
(RF
)
,
d
e
c
isio
n
tree
(DT)
,
a
n
d
su
p
p
o
rt
v
e
c
t
o
r
m
a
c
h
in
e
(S
V
M
)
m
o
d
e
ls.
T
h
e
m
o
st
fa
v
o
ra
b
le
o
u
tco
m
e
is
d
e
m
o
n
stra
ted
b
y
t
h
e
RF
m
o
d
e
l,
d
isp
lay
i
n
g
a
n
a
c
c
u
ra
c
y
o
f
9
3
%
.
In
t
h
is
m
a
n
n
e
r,
th
e
se
fin
d
in
g
s
m
a
y
b
e
h
e
lp
fu
l
f
o
r
in
jec
ti
o
n
m
o
l
d
i
n
g
b
u
si
n
e
ss
e
s
th
a
t
a
re
c
o
n
sid
e
ri
n
g
th
e
sig
n
ifi
c
a
n
c
e
o
f
u
sin
g
c
las
sifica
ti
o
n
a
l
g
o
rit
h
m
s
in
th
e
ir
m
a
n
u
fa
c
tu
rin
g
p
r
o
c
e
ss
.
K
ey
w
o
r
d
s
:
E
x
p
lain
ab
le
ar
tific
ial
in
tellig
e
n
ce
I
n
jectio
n
m
o
l
d
in
g
m
a
ch
in
e
L
ea
r
n
in
g
c
u
r
v
e
Ma
ch
in
e
lear
n
in
g
Ma
ch
in
e
s
elec
tio
n
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
:
Fao
u
zi
T
ay
alati
I
n
tellig
en
t A
u
to
m
atio
n
an
d
B
io
Me
d
Gen
o
m
ics
L
ab
o
r
ato
r
y
,
Facu
lty
o
f
Scien
ce
an
d
T
ec
h
n
o
lo
g
y
o
f
T
a
n
g
ier
Ab
d
elm
alek
E
s
s
aa
d
i U
n
iv
er
s
ity
T
eto
u
an
,
M
o
r
o
cc
o
E
m
ail:
f
ao
u
zi.
tay
alati@
etu
.
u
a
e.
ac
.
m
a
1.
I
NT
RO
D
UCT
I
O
N
Selectin
g
th
e
ap
p
r
o
p
r
iate
m
ac
h
in
e
f
o
r
a
n
ew
p
r
o
d
u
ct
r
em
ain
s
a
ch
allen
g
in
g
task
in
th
e
in
jectio
n
m
o
ld
in
g
in
d
u
s
tr
y
,
r
eq
u
i
r
in
g
e
x
ten
s
iv
e
k
n
o
wled
g
e
an
d
ex
p
e
r
ien
ce
[
1
]
.
T
h
e
d
ec
is
io
n
-
m
ak
i
n
g
p
r
o
ce
s
s
b
ec
o
m
es
in
cr
ea
s
in
g
ly
co
m
p
lex
in
s
m
al
l
an
d
m
ed
iu
m
-
s
ized
in
d
u
s
tr
ies,
esp
ec
ially
wh
en
m
a
n
ag
in
g
a
d
iv
er
s
e
r
a
n
g
e
o
f
p
r
o
d
u
cts
with
a
lim
ited
n
u
m
b
er
o
f
a
v
ailab
le
m
ac
h
in
es.
T
y
p
ically
,
d
eter
m
in
in
g
th
e
s
u
itab
le
m
ac
h
in
e
is
estab
lis
h
ed
d
u
r
in
g
th
e
in
itial
s
etu
p
th
r
o
u
g
h
th
eo
r
etica
l
ca
lc
u
latio
n
s
.
T
h
e
s
elec
tio
n
is
in
f
l
u
en
ce
d
b
y
v
ar
io
u
s
f
ac
to
r
s
,
in
clu
d
i
n
g
p
r
o
d
u
ct
ch
a
r
ac
ter
is
tics
s
u
ch
as
weig
h
t,
g
eo
m
etr
ic
s
p
ec
if
icatio
n
s
,
an
d
m
ater
ial
ty
p
e,
m
o
ld
s
p
ec
if
icatio
n
s
in
clu
d
in
g
weig
h
t
an
d
d
im
en
s
io
n
s
,
an
d
th
e
p
r
o
d
u
ctio
n
r
ate
r
eq
u
ested
b
y
th
e
cu
s
to
m
er
.
T
h
e
ch
o
s
en
m
ac
h
in
e
s
h
o
u
ld
m
ee
t
t
h
ese
f
u
n
d
am
e
n
tal
cr
iter
ia.
Ho
wev
er
,
d
e
v
iatio
n
s
f
r
o
m
th
eo
r
etica
l
ca
lcu
latio
n
s
ca
n
s
ig
n
if
ican
tly
im
p
ac
t
b
o
t
h
p
r
o
d
u
ct
m
ar
g
i
n
s
an
d
th
e
f
le
x
ib
ilit
y
b
etwe
en
m
ac
h
in
es.
Dis
cr
e
p
an
cies
i
n
th
e
co
s
t
ce
n
ter
s
o
f
m
ac
h
in
es
m
ea
n
th
at
s
elec
tin
g
th
e
wr
o
n
g
m
ac
h
i
n
e
ca
n
au
to
m
atica
lly
in
cr
ea
s
e
th
e
p
r
o
d
u
ct'
s
co
s
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Ma
ch
in
e
lea
r
n
in
g
-
a
s
s
is
ted
d
ec
is
io
n
s
u
p
p
o
r
t in
in
d
u
s
tr
ia
l .
.
.
(
F
a
o
u
z
i Ta
ya
la
ti)
271
b
r
ea
k
d
o
wn
[
2
]
.
T
h
er
e
f
o
r
e,
co
n
s
id
er
in
g
t
h
e
im
p
o
r
tan
ce
o
f
t
h
is
asp
ec
t,
it
is
cr
u
cial
to
ex
p
l
o
r
e
ap
p
r
o
ac
h
es
th
at
ca
n
h
elp
d
ec
is
io
n
-
m
ak
er
s
ac
c
u
r
ately
ass
ig
n
p
r
o
d
u
cts to
th
e
s
u
itab
le
m
ac
h
in
e.
I
n
th
is
s
tu
d
y
,
th
e
m
ain
o
b
j
ec
tiv
e
is
to
in
v
esti
g
ate
t
h
e
ap
p
licatio
n
o
f
m
ac
h
in
e
lear
n
in
g
(
ML
)
class
if
icatio
n
m
eth
o
d
s
to
s
elec
t
th
e
s
u
itab
le
m
ac
h
in
er
y
f
o
r
ea
ch
s
p
ec
if
ic
p
r
o
d
u
ct.
Acc
o
r
d
in
g
to
th
e
liter
atu
r
e
,
m
ac
h
in
e
s
elec
tio
n
is
ca
teg
o
r
iz
ed
as
a
m
u
lti
-
cr
iter
ia
p
r
o
b
lem
[
3
]
.
Var
i
o
u
s
class
ical
m
eth
o
d
s
h
av
e
b
ee
n
e
m
p
lo
y
ed
to
ass
i
s
t
d
ec
is
io
n
-
m
ak
er
s
in
s
elec
tin
g
th
e
o
p
tim
al
ch
o
ice.
Am
o
n
g
th
e
p
r
o
m
in
en
t
m
eth
o
d
s
ar
e
th
e
an
aly
tical
h
ier
ar
ch
y
p
r
o
ce
s
s
(
AHP)
[
4
]
,
[
5
]
an
d
th
e
tech
n
iq
u
e
f
o
r
o
r
d
er
p
r
ef
er
en
ce
b
y
s
im
ilar
ity
to
id
ea
l
s
o
lu
tio
n
(
T
OPSIS)
[
6
]
.
T
h
e
AHP
f
u
n
ctio
n
s
as
a
f
r
am
ewo
r
k
f
o
r
s
tr
u
ct
u
r
in
g
an
d
an
aly
zi
n
g
co
m
p
lex
d
e
cisi
o
n
s
,
in
teg
r
atin
g
m
ath
em
atica
l
p
r
in
cip
les
an
d
p
s
y
ch
o
lo
g
ical
f
ac
t
o
r
s
.
I
t
e
n
ab
le
s
th
e
ev
alu
atio
n
o
f
t
h
ese
elem
en
ts
co
n
ce
r
n
i
n
g
th
e
o
v
er
ar
c
h
in
g
o
b
jectiv
e,
th
er
eb
y
f
ac
ilit
atin
g
a
s
y
s
tem
atic
d
ec
is
io
n
-
m
ak
in
g
ap
p
r
o
ac
h
.
Me
an
wh
ile,
th
e
T
OPSIS
m
eth
o
d
o
p
er
ates
o
n
th
e
p
r
in
c
ip
le
o
f
s
elec
tin
g
an
alter
n
ativ
e
th
at
m
in
im
izes
th
e
g
eo
m
etr
ic
d
is
tan
ce
to
th
e
p
o
s
itiv
e
id
ea
l so
lu
tio
n
wh
ile
m
ax
im
izin
g
th
e
g
eo
m
etr
ic
d
is
tan
ce
f
r
o
m
th
e
n
e
g
ativ
e
id
ea
l so
lu
tio
n
.
Alter
n
ativ
es
ar
e
r
an
k
ed
th
r
o
u
g
h
th
e
ca
lcu
la
tio
n
o
f
a
n
o
v
er
all
in
d
ex
b
ased
o
n
th
eir
d
is
tan
ce
s
f
r
o
m
th
ese
id
ea
l
s
o
lu
tio
n
s
[
7
]
.
B
o
th
m
eth
o
d
s
in
v
o
l
v
e
s
u
b
ject
iv
e
d
ec
is
io
n
s
in
r
an
k
in
g
cr
iter
ia
an
d
ch
o
ices,
in
tr
o
d
u
cin
g
a
p
o
s
s
ib
le
s
o
u
r
ce
o
f
b
ias in
to
th
e
d
ec
is
io
n
-
m
a
k
in
g
p
r
o
ce
s
s
.
T
h
er
ef
o
r
e,
th
is
p
ap
e
r
ex
p
l
o
r
es
an
alter
n
ativ
e
ap
p
r
o
ac
h
b
ased
o
n
ML
to
ai
d
in
s
elec
tin
g
th
e
a
p
p
r
o
p
r
iate
in
jectio
n
-
m
o
ld
in
g
m
ac
h
in
e
f
o
r
a
g
iv
en
p
r
o
d
u
ct.
Six
class
if
icatio
n
ML
alg
o
r
ith
m
s
ar
e
u
tili
ze
d
,
in
clu
d
in
g
th
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM
)
,
r
an
d
o
m
f
o
r
est
(
R
F)
,
d
ec
is
io
n
tr
ee
(
DT
)
,
n
aiv
e
B
ay
es
(
NB
)
m
o
d
el,
s
to
ch
asti
c
g
r
ad
ien
t
d
escen
t
(
SGD
)
c
lass
if
ier
,
an
d
lo
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
.
E
v
alu
atio
n
m
etr
ics
s
u
ch
as
ac
cu
r
ac
y
,
r
ec
all,
F1
s
co
r
e,
an
d
p
r
ec
is
io
n
ar
e
e
m
p
l
o
y
ed
.
T
h
e
s
tu
d
y
'
s
f
in
d
in
g
s
s
u
g
g
est
th
at
th
ese
class
if
i
ca
tio
n
m
eth
o
d
s
ef
f
e
ctiv
ely
ca
p
tu
r
e
f
ea
tu
r
e
r
elatio
n
s
h
ip
s
,
en
ab
lin
g
p
r
ed
ictio
n
s
o
f
m
ac
h
i
n
e
s
elec
tio
n
.
T
h
e
RF
,
DT
,
an
d
SVM
m
o
d
els
ca
n
p
r
ed
ict
th
e
co
r
r
ec
t
m
ac
h
i
n
e
s
elec
tio
n
with
ac
cu
r
ac
ies
o
f
ap
p
r
o
x
im
ately
9
3
%,
8
7
%,
ev
en
with
lim
ited
tr
ain
in
g
d
ata.
T
h
ese
p
r
o
m
is
in
g
r
esu
lts
ca
n
p
r
o
v
id
e
v
al
u
ab
le
g
u
id
an
ce
f
o
r
s
elec
tin
g
n
ew
m
ac
h
in
es f
o
r
u
p
co
m
in
g
p
ar
ts
.
T
h
e
ar
ticle
is
d
iv
id
ed
in
to
f
o
u
r
m
ain
s
ec
tio
n
s
.
T
h
e
f
ir
s
t
s
ec
tio
n
in
tr
o
d
u
ce
s
th
e
in
jectio
n
m
o
ld
in
g
p
r
o
ce
s
s
,
em
p
h
asizin
g
th
e
co
n
v
en
tio
n
al
ap
p
r
o
ac
h
to
s
elec
tin
g
m
ac
h
in
es
f
o
r
n
ew
p
r
o
d
u
cts.
T
h
e
s
ec
o
n
d
s
ec
tio
n
r
ev
iews
ex
is
tin
g
liter
atu
r
e
o
n
m
ac
h
in
e
s
elec
tio
n
p
r
o
ce
s
s
es.
T
h
e
th
ir
d
s
ec
tio
n
d
is
cu
s
s
es
th
e
a
p
p
licatio
n
o
f
class
if
icatio
n
m
eth
o
d
s
u
s
in
g
d
ata
g
ath
er
ed
f
r
o
m
r
ea
l
in
d
u
s
tr
y
s
itu
atio
n
s
.
L
astl
y
,
th
e
f
o
u
r
th
s
ec
tio
n
d
elv
es
in
to
th
e
r
esu
lts
o
b
tain
ed
an
d
co
m
p
ar
es th
e
p
er
f
o
r
m
an
ce
o
f
th
e
m
eth
o
d
s
.
2.
SCO
P
E
O
F
ST
U
DY
AND
L
I
T
E
RA
T
UR
E
RE
V
I
E
W
2
.
1
.
I
nje
ct
io
n
m
o
ldi
n pro
ce
s
s
T
h
e
in
jectio
n
p
r
o
ce
s
s
tech
n
iq
u
e
r
elies o
n
th
e
p
r
in
cip
le
o
f
tr
a
n
s
f
o
r
m
in
g
p
o
ly
m
e
r
s
f
r
o
m
a
s
o
lid
s
tate
to
a
m
o
lten
s
tate
b
y
u
s
in
g
a
p
r
e
s
s
m
ac
h
in
e
an
d
s
p
ec
if
ic
to
o
lin
g
[
8
]
,
[
9
]
.
T
h
e
in
jectio
n
m
ac
h
in
e
co
m
p
r
is
es
two
p
r
im
ar
y
co
m
p
o
n
e
n
ts
:
th
e
in
jec
tio
n
u
n
it
an
d
th
e
clo
s
in
g
u
n
it.
T
h
e
in
jectio
n
u
n
it
is
r
esp
o
n
s
ib
le
f
o
r
co
n
v
er
tin
g
th
e
p
last
ic
f
r
o
m
a
s
o
lid
s
tate
to
a
p
last
icize
d
s
tate
b
ef
o
r
e
in
jecti
n
g
it
in
to
th
e
m
o
ld
.
Me
an
wh
il
e,
th
e
clo
s
in
g
u
n
it
is
r
esp
o
n
s
ib
le
f
o
r
s
ea
lin
g
th
e
to
o
lin
g
,
lo
ck
in
g
it
in
p
lace
,
an
d
o
v
er
s
ee
in
g
th
e
co
o
lin
g
p
r
o
ce
s
s
o
f
th
e
p
last
ic.
T
h
e
p
r
o
d
u
ctio
n
cy
cle
ca
n
b
e
r
ep
r
es
en
ted
b
y
f
iv
e
m
ain
s
eq
u
en
ce
s
,
as illu
s
tr
ated
in
F
ig
u
r
e
1
.
Fig
u
r
e
1
.
T
h
e
m
ain
s
tep
s
o
f
th
e
in
jectio
n
m
o
ld
i
n
g
p
r
o
ce
s
s
Du
r
in
g
th
e
d
o
s
in
g
an
d
p
last
i
cizin
g
p
h
ase,
th
e
p
last
ic
m
ater
ial
f
alls
d
o
wn
f
r
o
m
th
e
h
o
p
p
er
t
o
th
e
in
jectio
n
s
cr
ew
u
n
d
er
t
h
e
in
f
lu
en
ce
o
f
g
r
av
ity
.
Sev
er
al
m
ater
ials
ca
n
b
e
m
o
ld
ed
,
s
o
m
e
wid
ely
u
s
er
ar
e
p
o
ly
am
id
e
(
PA
6
)
,
p
o
ly
ester
,
p
o
ly
v
in
y
l
c
h
lo
r
id
e
,
n
y
lo
n
,
an
d
a
cr
y
lic
[
1
0
]
.
Su
b
s
eq
u
en
tly
,
th
e
p
o
ly
m
er
u
n
d
er
g
o
es
tr
an
s
f
o
r
m
atio
n
i
n
to
a
m
o
lten
s
tate
th
r
o
u
g
h
a
d
u
al
m
ec
h
an
is
m
:
th
e
s
h
ea
r
in
g
ef
f
ec
t
i
n
d
u
ce
d
b
y
th
e
s
cr
ew
an
d
th
e
h
ea
tin
g
ef
f
ec
t
g
en
e
r
ated
b
y
h
e
atin
g
r
esis
tan
ce
s
s
u
r
r
o
u
n
d
in
g
t
h
e
s
cr
ew.
T
h
e
m
o
lten
p
o
ly
m
er
is
th
en
cy
cled
b
ac
k
in
f
r
o
n
t
o
f
th
e
s
cr
ew
b
y
m
et
er
in
g
th
e
to
tal
v
o
l
u
m
e
f
o
r
a
co
m
p
lete
p
last
ic
p
ar
t
c
y
cle.
I
n
th
e
f
illi
n
g
p
h
ase,
co
m
m
o
n
l
y
k
n
o
wn
as
th
e
i
n
jectio
n
p
h
ase,
th
e
p
last
ic
is
in
tr
o
d
u
ce
d
in
t
o
th
e
m
o
ld
th
r
o
u
g
h
f
ill
in
g
ch
a
n
n
els
b
y
th
e
s
cr
ew.
T
h
is
p
r
o
c
ess
ty
p
ically
l
asts
a
f
ew
s
ec
o
n
d
s
,
with
th
e
d
u
r
atio
n
i
n
f
lu
en
ce
d
b
y
th
e
m
at
er
ial'
s
v
is
co
s
ity
an
d
th
e
d
esire
d
in
jectio
n
f
lo
w
r
ate.
T
h
e
in
jectio
n
p
r
ess
u
r
e
d
u
r
i
n
g
th
is
p
h
ase
ca
n
r
a
n
g
e
f
r
o
m
a
f
ew
h
u
n
d
r
ed
to
a
th
o
u
s
an
d
b
ar
s
.
Du
r
i
n
g
th
e
p
ac
k
in
g
p
h
ase,
th
e
s
cr
ew
m
ain
tain
s
p
r
ess
u
r
e
o
n
t
h
e
p
last
ic
to
a
d
d
r
ess
p
o
ten
tial
is
s
u
es
ar
is
in
g
f
r
o
m
m
ater
ial
s
h
r
in
k
a
g
e
as
it
s
tar
ts
to
co
o
l.
T
h
is
ad
d
itio
n
al
p
r
ess
u
r
e
co
m
p
en
s
ates
f
o
r
an
y
p
r
o
b
lem
s
th
at
m
ay
ar
is
e
d
u
r
in
g
th
e
co
o
lin
g
p
r
o
ce
s
s
.
I
n
t
h
e
f
in
al
s
tep
,
th
e
co
o
lin
g
p
h
ase
aim
s
to
s
o
lid
if
y
th
e
en
tir
e
p
ar
t
u
n
til
th
e
d
esire
d
r
ig
id
ity
is
ac
h
iev
e
d
.
T
h
e
s
o
lid
if
icatio
n
p
r
o
ce
s
s
is
f
ac
ilit
ated
b
y
th
e
th
er
m
al
tr
a
n
s
f
er
ap
p
lied
b
y
th
e
m
o
ld
.
I
n
th
e
ejec
tio
n
p
h
ase,
a
f
t
er
th
e
p
ar
t
h
as
attain
ed
s
u
f
f
icien
t
r
ig
id
ity
,
th
e
m
o
ld
o
p
e
n
s
,
an
d
th
e
p
ar
t
is
ejec
ted
D
o
z
i
n
g
a
n
d
p
l
a
st
i
c
i
z
i
n
g
p
h
a
se
F
i
l
i
n
g
p
h
a
se
P
a
c
k
i
n
g
p
h
a
se
C
o
o
l
i
n
g
p
h
a
se
Ej
e
c
t
i
o
n
p
h
a
se
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
270
-
2
8
5
272
th
r
o
u
g
h
a
d
esig
n
ated
ejec
tio
n
m
ec
h
an
is
m
.
E
ac
h
s
tep
o
f
th
e
p
r
o
ce
s
s
is
ca
r
ef
u
lly
d
ef
i
n
ed
b
ased
o
n
s
p
ec
if
ic
p
ar
am
eter
s
.
2
.
2
.
M
a
chine selec
t
io
n:
co
nv
ent
io
na
l a
pp
ro
a
ch
Selectin
g
th
e
r
ig
h
t
m
ac
h
in
e
f
o
r
a
p
ar
tic
u
lar
p
r
o
d
u
ct
r
e
m
ain
s
a
task
d
em
a
n
d
in
g
ex
p
er
tis
e
an
d
ex
p
er
ien
ce
.
T
h
e
m
o
ld
in
g
ad
ju
s
ter
allo
ca
tes
m
ac
h
in
es
b
ased
o
n
v
ar
io
u
s
cr
iter
ia,
in
clu
d
in
g
p
ar
t
d
esig
n
,
m
o
l
d
d
esig
n
,
m
ater
ial
s
p
ec
if
icatio
n
s
,
an
d
cu
s
to
m
er
d
e
m
an
d
[
1
1
]
.
Firstl
y
,
m
o
ld
c
h
ar
ac
te
r
is
tics
,
p
ar
ticu
lar
ly
d
im
en
s
io
n
s
s
u
ch
as
th
ick
n
ess
,
wid
th
,
len
g
th
,
h
eig
h
t,
a
n
d
weig
h
t,
ar
e
p
iv
o
tal
in
ass
ess
i
n
g
th
e
f
ea
s
ib
ilit
y
o
f
m
o
u
n
tin
g
th
e
to
o
l
o
n
t
h
e
m
ac
h
in
e.
I
f
t
h
e
m
o
ld
s
u
r
p
ass
es
th
e
s
p
ec
if
ied
d
im
en
s
io
n
s
o
f
th
e
m
ac
h
in
e
,
it
m
ig
h
t
r
eq
u
ir
e
allo
ca
tio
n
to
a
lar
g
er
-
s
ized
p
r
ess
to
ac
co
m
m
o
d
ate
i
ts
s
ize
ad
eq
u
ately
.
Seco
n
d
ly
,
p
ar
t
ch
ar
ac
ter
is
tics
,
en
co
m
p
ass
in
g
d
im
en
s
io
n
s
,
we
ig
h
t,
an
d
th
ick
n
ess
,
co
n
s
titu
te
th
e
m
o
s
t
s
ig
n
if
ican
t
p
a
r
am
et
er
s
[
1
2
]
.
T
h
e
p
ar
t'
s
d
im
en
s
io
n
s
d
ir
ec
tly
im
p
ac
t
th
e
m
ac
h
in
e'
s
o
p
en
in
g
s
tr
o
k
e
,
wh
ich
is
cr
itical
f
o
r
s
u
cc
ess
f
u
l
p
ar
t
ejec
tio
n
af
ter
s
o
lid
if
icatio
n
.
I
f
th
e
o
p
en
in
g
s
tr
o
k
e
is
in
ad
eq
u
ate
f
o
r
ejec
tio
n
,
s
elec
tin
g
a
m
ac
h
in
e
wit
h
a
lar
g
er
o
p
en
in
g
ca
p
ac
ity
m
ay
b
e
n
ec
ess
ar
y
.
Fu
r
th
er
m
o
r
e
,
th
e
p
ar
t'
s
th
ick
n
ess
d
ir
ec
tly
af
f
e
cts
th
e
m
ac
h
in
e'
s
m
ax
im
u
m
in
jectio
n
p
r
ess
u
r
e
ca
p
ac
ity
;
th
in
n
er
p
a
r
ts
n
ec
ess
ita
te
h
ig
h
er
in
jectio
n
p
r
ess
u
r
e.
T
h
ir
d
ly
,
m
ater
ial
ch
ar
ac
ter
is
tics
ar
e
p
iv
o
tal
in
th
e
p
r
o
ce
s
s
,
en
co
m
p
ass
in
g
v
ar
io
u
s
ty
p
es
with
d
iv
er
s
e
ch
em
ical
co
m
p
o
s
itio
n
s
[
1
3
]
.
Fro
m
s
im
p
le
to
h
ig
h
ly
tech
n
ical
an
d
ag
g
r
ess
iv
e
m
ater
ials
,
th
e
ch
em
ical
co
m
p
o
s
itio
n
s
ig
n
if
ican
tly
im
p
ac
ts
th
e
s
elec
tio
n
o
f
th
e
in
jectio
n
u
n
it.
T
h
e
m
ac
h
in
e
s
elec
tio
n
p
r
o
ce
s
s
m
u
s
t
ac
co
u
n
t
f
o
r
th
e
s
p
ec
if
icatio
n
s
o
f
th
e
s
cr
ew
an
d
b
ar
r
els,
esp
ec
ially
th
eir
tr
e
atm
en
t
to
with
s
tan
d
th
e
ch
em
ical
ag
g
r
es
s
iv
en
ess
o
f
th
e
m
ater
ial
b
ein
g
p
r
o
ce
s
s
ed
.
Fin
ally
,
th
e
p
r
o
d
u
ctio
n
r
ate
s
ig
n
if
ies
th
e
ca
p
ac
ity
o
f
th
e
c
h
o
s
en
m
ac
h
in
e
to
m
ee
t
s
p
ec
if
ic
r
ates
alig
n
ed
with
cu
s
to
m
er
d
em
an
d
s
.
T
h
is
f
ac
to
r
f
ac
ilit
ates
th
e
s
izin
g
o
f
th
e
in
jec
tio
n
u
n
it,
en
s
u
r
in
g
its
ca
p
ac
ity
alig
n
s
with
th
e
r
eq
u
ir
ed
p
r
o
d
u
ctio
n
r
ate.
T
h
e
m
ac
h
in
e
s
elec
tio
n
p
r
o
ce
s
s
ca
n
b
e
illu
s
tr
ated
b
y
Fig
u
r
e
2
.
I
t
ca
n
b
e
s
eg
m
en
ted
in
to
t
h
r
ee
p
r
im
ar
y
p
h
ases
: f
o
r
ce
ca
lcu
lat
io
n
,
clam
p
in
g
u
n
it v
alid
atio
n
,
an
d
in
jectio
n
u
n
it v
alid
atio
n
.
Fig
u
r
e
2
.
Ma
ch
i
n
e
s
elec
tio
n
f
l
o
wh
ch
ar
t: c
o
n
v
en
tio
n
al
ap
p
r
o
ac
h
I
n
th
e
in
itial
p
h
ase,
th
e
f
o
r
ce
c
alcu
latio
n
in
v
o
l
v
es
d
eter
m
in
in
g
th
e
clam
p
i
n
g
f
o
r
ce
r
e
q
u
ir
ed
to
en
s
u
r
e
th
e
p
r
o
p
er
clo
s
in
g
an
d
h
o
l
d
in
g
o
f
th
e
m
o
ld
d
u
r
in
g
th
e
p
las
tic
in
jectio
n
,
p
ac
k
in
g
,
an
d
co
o
lin
g
p
h
ases
.
I
f
th
e
ca
lcu
lated
f
o
r
ce
is
with
in
th
e
m
ax
im
u
m
clam
p
in
g
f
o
r
ce
ca
p
ac
ity
o
f
th
e
m
ac
h
in
e,
th
e
m
ac
h
in
e
ca
n
b
e
v
alid
ated
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Ma
ch
in
e
lea
r
n
in
g
-
a
s
s
is
ted
d
ec
is
io
n
s
u
p
p
o
r
t in
in
d
u
s
tr
ia
l .
.
.
(
F
a
o
u
z
i Ta
ya
la
ti)
273
b
ased
o
n
th
is
f
o
r
ce
,
allo
win
g
th
e
p
r
o
ce
s
s
to
p
r
o
ce
ed
to
th
e
s
e
co
n
d
s
tep
.
Ho
wev
er
,
if
th
e
ca
lc
u
lated
f
o
r
ce
ex
ce
e
d
s
th
e
m
ac
h
in
e'
s
m
ax
im
u
m
clam
p
in
g
f
o
r
ce
,
it
b
ec
o
m
es
n
ec
ess
ar
y
to
co
n
s
id
er
a
m
ac
h
in
e
wit
h
a
h
ig
h
er
clo
s
in
g
ca
p
ac
ity
,
an
d
f
u
r
th
e
r
s
tep
s
m
ay
n
o
t
b
e
p
u
r
s
u
ed
.
I
n
th
e
s
ec
o
n
d
p
h
ase,
v
alid
atin
g
th
e
cla
m
p
in
g
u
n
it
in
clu
d
es
co
n
f
ir
m
in
g
wh
eth
er
th
e
m
o
ld
c
an
b
e
ef
f
ec
tiv
ely
m
o
u
n
ted
o
n
t
h
e
m
ac
h
in
e,
co
n
s
id
er
i
n
g
its
s
iz
e
an
d
th
e
allo
wab
le
s
p
ac
es
with
in
th
e
m
ac
h
in
e
.
Ad
d
itio
n
ally
,
th
e
v
alid
atio
n
c
o
n
s
id
er
s
also
th
e
v
e
r
if
icatio
n
o
f
th
e
m
ec
h
a
n
ical
d
im
en
s
io
n
s
o
f
th
e
m
o
ld
a
n
d
th
e
o
p
en
in
g
s
tr
o
k
e
allo
win
g
s
a
tis
f
ac
to
r
y
ejec
tio
n
o
f
th
e
p
ar
t
d
u
r
in
g
th
e
ejec
tio
n
p
h
ase.
I
f
th
ese
cr
iter
ia
ar
e
m
et,
th
e
clam
p
in
g
u
n
it
is
v
alid
ated
.
Ho
wev
er
,
if
th
e
r
eq
u
ir
em
e
n
ts
ar
e
n
o
t
m
et,
an
o
v
er
s
ize
d
m
ac
h
in
e
m
a
y
b
e
n
ec
ess
ar
y
,
an
d
th
e
s
elec
tio
n
p
r
o
ce
s
s
m
ay
n
ee
d
to
b
e
ig
n
o
r
ed
.
I
n
s
u
ch
ca
s
es,
co
n
s
id
er
atio
n
o
f
a
d
if
f
er
en
t
p
r
ess
b
ec
o
m
es
n
ec
ess
ar
y
.
T
h
e
t
h
ir
d
an
d
f
in
al
p
h
ase
is
d
ed
icate
d
to
v
alid
atin
g
th
e
in
jectio
n
u
n
it.
T
h
e
u
n
it
is
co
n
s
id
er
ed
s
u
itab
le
if
i
t
m
ee
ts
th
e
c
r
iter
ia
r
elate
d
to
th
e
m
ater
ial,
d
o
s
in
g
v
o
lu
m
e,
p
ar
t
weig
h
t,
ty
p
e
o
f
s
cr
ew,
m
a
x
im
u
m
in
jectio
n
p
r
ess
u
r
e,
a
n
d
m
a
x
im
u
m
p
r
o
d
u
ctio
n
r
ate.
I
f
a
n
y
o
f
th
ese
cr
iter
ia
is
n
o
t
m
et,
th
e
allo
ca
tio
n
o
f
th
e
p
ar
t
ca
n
n
o
t
b
e
ass
u
r
ed
,
an
d
f
u
r
t
h
er
co
n
s
id
er
atio
n
o
r
a
d
ju
s
tm
en
ts
m
ay
b
e
r
eq
u
ir
ed
.
2
.
3
.
Rela
t
ed
wo
rk
s
Ma
n
y
W
h
ile
n
u
m
er
o
u
s
s
tu
d
ies
h
av
e
d
el
v
ed
in
to
m
ac
h
in
e
s
elec
tio
n
allo
ca
tio
n
,
o
n
ly
a
lim
ited
f
ew
h
av
e
tack
led
th
e
is
s
u
e
u
s
in
g
ar
tific
ial
in
tellig
en
ce
m
et
h
o
d
s
.
I
n
an
in
v
esti
g
atio
n
co
n
d
u
cte
d
b
y
Ny
an
g
a
et
a
l
.
[
1
4
]
,
t
h
e
s
tu
d
y
ex
p
lo
r
ed
th
e
ap
p
licatio
n
o
f
a
m
u
lti
-
ag
e
n
t
s
y
s
tem
f
o
r
s
elec
tin
g
th
e
m
o
s
t
s
u
itab
le
m
ac
h
in
e
t
o
o
p
tim
ize
p
r
o
d
u
ctio
n
o
r
d
er
.
T
h
e
c
r
iter
ia
f
o
r
m
ac
h
in
e
s
elec
tio
n
wer
e
es
tab
lis
h
ed
u
s
in
g
th
e
AHP,
with
q
u
ality
,
tim
e,
an
d
co
s
t
id
en
tifie
d
as
k
ey
p
ar
am
et
er
s
.
Fiv
e
cr
iter
ia
—
clam
p
in
g
f
o
r
ce
,
d
is
tan
ce
s
b
etwe
en
tie
b
ar
s
,
s
h
o
t
weig
h
t,
tim
e,
an
d
co
s
t
—
wer
e
ch
o
s
en
.
A
m
u
lti
-
ag
en
t
s
y
s
tem
,
co
m
p
r
is
in
g
a
m
an
ag
in
g
ag
en
t,
a
d
m
in
is
tr
ato
r
,
an
d
b
id
d
in
g
ag
en
ts
,
f
ac
ilit
ated
an
au
ctio
n
b
etwe
en
m
ac
h
in
e
ag
en
ts
ac
tin
g
as
p
o
ten
tial
co
n
t
r
ac
to
r
s
r
ep
r
esen
tin
g
av
ailab
le
m
ac
h
in
es
in
th
e
p
r
o
d
u
ctio
n
lin
e.
T
h
e
s
tu
d
y
d
e
m
o
n
s
tr
ated
t
h
at
th
e
s
y
s
tem
ef
f
ec
tiv
ely
r
ed
u
ce
d
s
h
o
r
t
s
h
o
t
d
ef
ec
ts
b
y
1
0
%,
s
in
k
m
ar
k
s
b
y
5
%,
a
n
d
f
lash
in
g
b
y
3
0
%.
T
r
iv
e
d
i
et
a
l.
[
1
5
]
p
r
esen
ted
a
n
ap
p
licatio
n
o
f
m
u
lti
-
cr
iter
ia
d
ec
is
io
n
-
m
ak
in
g
u
s
in
g
th
e
o
p
t
im
ized
T
OPSIS
m
eth
o
d
with
a
f
u
zz
y
lo
g
ic
a
p
p
r
o
ac
h
.
T
h
is
r
esear
ch
f
o
cu
s
ed
o
n
1
4
elec
tr
ical
m
o
ld
in
g
m
ac
h
i
n
es,
co
n
s
id
er
in
g
eig
h
t
cr
iter
i
a:
r
ate
o
f
r
etu
r
n
,
r
is
k
in
v
estme
n
t,
lik
ely
p
r
o
f
it,
in
s
tallatio
n
co
s
t,
s
im
ilar
ity
to
ex
is
tin
g
b
u
s
in
ess
,
en
v
ir
o
n
m
e
n
tal
im
p
ac
t,
an
d
ex
p
ec
te
d
life
.
B
y
in
co
r
p
o
r
atin
g
f
u
zz
y
lo
g
ic,
th
e
f
u
zz
y
T
OPSIS
m
eth
o
d
p
r
o
v
id
e
d
r
esu
lts
o
r
ien
ted
to
war
d
s
elec
ted
m
e
m
b
er
s
h
ip
f
u
n
ctio
n
s
,
th
er
eb
y
m
a
k
in
g
th
e
m
ac
h
in
e
s
elec
tio
n
p
r
o
ce
s
s
m
o
r
e
au
to
n
o
m
o
u
s
.
L
i
n
an
d
Ya
n
g
[
5
]
d
ev
elo
p
e
d
a
m
o
d
el
em
p
lo
y
in
g
th
e
AHP
to
s
e
lect
th
e
m
o
s
t
ap
p
r
o
p
r
iate
m
ac
h
in
e
.
T
h
eir
ap
p
r
o
ac
h
ce
n
ter
ed
o
n
c
h
o
o
s
in
g
a
m
ac
h
in
e
f
r
o
m
v
ar
io
u
s
av
ailab
le
o
p
tio
n
s
tailo
r
ed
f
o
r
m
a
n
u
f
ac
tu
r
in
g
s
p
ec
if
ic
ty
p
es
o
f
p
ar
ts
.
T
h
e
r
esea
r
ch
er
s
in
tr
o
d
u
ce
d
a
p
r
o
to
ty
p
e
f
r
am
ewo
r
k
a
n
d
s
o
f
twar
e
th
at
u
tili
ze
d
AHP
f
o
r
m
ac
h
in
e
s
elec
tio
n
,
in
teg
r
atin
g
th
e
ex
p
e
r
t
-
s
y
s
tem
co
n
ce
p
t
t
o
d
ev
elo
p
th
e
p
r
o
g
r
a
m
an
d
ar
ticu
late
s
elec
tio
n
cr
it
er
ia.
T
h
eir
p
r
im
ar
y
g
o
al
was
t
o
o
p
tim
ize
m
ac
h
i
n
e
s
elec
tio
n
,
en
h
an
cin
g
b
o
th
co
s
t
-
ef
f
ec
tiv
en
ess
an
d
wo
r
k
s
h
o
p
f
lex
ib
ilit
y
with
ex
is
tin
g
p
ar
ts
.
I
n
a
r
ec
en
t
s
tu
d
y
b
y
Do
m
in
g
u
ez
et
a
l.
[
1
6
]
,
m
u
lt
icr
iter
ia
s
elec
tio
n
m
eth
o
d
s
s
u
ch
as
co
m
b
i
n
ativ
e
d
is
tan
ce
-
b
ased
ass
ess
m
en
t
(
C
ODAS
)
,
T
OPSIS,
an
d
AHP
wer
e
in
v
esti
g
ated
in
th
e
p
astry
in
d
u
s
tr
y
.
T
h
e
s
tu
d
y
ai
m
ed
to
c
o
m
p
ar
e
th
e
p
er
f
o
r
m
an
ce
o
f
ea
ch
m
eth
o
d
in
m
ak
in
g
th
e
b
est
ch
o
ice
am
o
n
g
f
iv
e
m
ac
h
in
es.
C
r
iter
ia
co
n
s
id
er
ed
in
clu
d
ed
p
r
ice,
ca
p
ac
ity
,
war
r
an
ty
,
p
r
o
d
u
ct
weig
h
t,
an
d
s
p
ee
d
.
R
e
s
u
lts
in
d
icate
d
th
at
AH
P
an
d
C
ODA
S
p
r
o
v
id
ed
ap
p
r
o
x
im
ate
m
ac
h
in
e
s
elec
tio
n
s
,
with
p
o
ten
tial
r
ef
in
em
en
t
th
r
o
u
g
h
co
m
p
lem
e
n
tar
y
s
tu
d
ies
u
s
in
g
v
ar
ian
ce
an
aly
s
is
.
T
h
is
s
tu
d
y
h
i
g
h
lig
h
te
d
th
e
e
f
f
ec
tiv
en
ess
o
f
th
ese
m
u
lticr
iter
ia
s
elec
tio
n
m
eth
o
d
s
in
ch
o
o
s
in
g
m
ac
h
in
es
f
o
r
th
e
p
astry
i
n
d
u
s
tr
y
.
ML
-
a
s
s
is
ted
d
ec
is
io
n
s
u
p
p
o
r
t
in
in
d
u
s
tr
ial
m
an
u
f
ac
tu
r
in
g
ca
n
also
b
e
co
n
s
id
er
ed
i
n
r
elatio
n
with
d
ee
p
lear
n
in
g
-
as
s
is
ted
s
m
ar
t
p
r
o
ce
s
s
p
lan
n
in
g
,
r
o
b
o
tic
wir
eless
s
en
s
o
r
n
etwo
r
k
s
,
an
d
g
e
o
s
p
atial
b
ig
d
ata
m
a
n
ag
em
en
t
alg
o
r
ith
m
s
with
in
th
e
d
o
m
ain
o
f
th
e
in
ter
n
et
o
f
m
an
u
f
ac
tu
r
in
g
th
i
n
g
s
,
as
d
is
cu
s
s
ed
in
[
1
7
]
.
Ad
d
itio
n
ally
,
it
r
elate
s
to
ar
tific
ial
in
tellig
en
ce
-
b
ased
d
ec
is
io
n
-
m
ak
in
g
alg
o
r
ith
m
s
,
in
ter
n
et
o
f
th
in
g
s
s
en
s
in
g
n
etwo
r
k
s
,
an
d
s
u
s
tain
ab
le
cy
b
er
-
p
h
y
s
ical
m
a
n
a
g
em
en
t
s
y
s
tem
s
in
b
ig
d
at
a
-
d
r
iv
en
co
g
n
itiv
e
m
an
u
f
ac
tu
r
in
g
,
as o
u
tlin
ed
in
[
1
8
]
.
3.
I
M
P
L
E
M
E
NT
A
T
I
O
N
3
.
1
.
M
et
ho
do
lo
g
y
T
h
e
s
u
g
g
ested
ap
p
r
o
ac
h
is
b
ased
o
n
th
e
tr
ad
itio
n
al
co
n
ce
p
t
o
f
ML
,
it
r
ep
r
esen
ted
o
n
th
e
F
ig
u
r
e
3
.
T
h
e
m
eth
o
d
o
l
o
g
y
c
o
m
p
r
is
es
f
iv
e
p
r
im
ar
y
s
tep
s
.
I
n
itially
,
d
ata
co
llectio
n
is
co
n
d
u
cted
to
g
ath
er
th
e
r
e
q
u
is
ite
in
f
o
r
m
atio
n
f
o
r
m
o
d
el
ap
p
licatio
n
.
Fo
llo
win
g
t
h
is
,
th
e
s
ec
o
n
d
s
tep
in
v
o
lv
es
p
r
e
p
r
o
ce
s
s
in
g
th
e
d
ata,
aim
in
g
t
o
ascer
tain
co
r
r
elatio
n
s
an
d
r
ea
d
y
it
f
o
r
m
o
d
el
im
p
lem
e
n
tatio
n
.
T
h
e
s
u
b
s
eq
u
en
t
s
tep
en
tai
ls
th
e
ap
p
licatio
n
o
f
ML
alg
o
r
ith
m
s
.
I
n
th
e
f
o
u
r
t
h
s
tep
,
p
er
f
o
r
m
a
n
ce
co
m
p
ar
is
o
n
am
o
n
g
d
i
f
f
er
en
t
m
et
h
o
d
s
is
ca
r
r
ied
o
u
t
to
s
elec
t
th
e
o
p
tim
al
ap
p
r
o
ac
h
.
L
astl
y
,
th
e
f
if
th
s
tep
in
v
o
lv
es
an
aly
s
in
g
th
e
lear
n
in
g
p
r
o
ce
s
s
o
f
ea
ch
m
o
d
el
f
r
o
m
th
e
p
r
o
v
id
e
d
d
ata.
Ad
d
itio
n
ally
,
a
f
in
al
v
er
if
icatio
n
is
p
er
f
o
r
m
e
d
to
ass
ess
th
e
in
f
lu
en
ce
o
f
i
n
d
iv
id
u
al
f
ea
tu
r
es
o
n
th
e
m
o
d
els u
s
in
g
t
h
e
lo
ca
l in
te
r
p
r
etab
le
m
o
d
el
-
ag
n
o
s
tic
ex
p
l
an
atio
n
s
(
L
I
ME
)
m
eth
o
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
270
-
2
8
5
274
Fig
u
r
e
3
.
Flo
w
ch
a
r
t o
f
t
h
e
s
u
g
g
ested
ap
r
o
ac
h
3
.
2
.
I
nitia
l da
t
a
T
h
e
s
tu
d
y
u
s
ed
d
ataset
co
lle
cted
f
r
o
m
ac
tu
al
wo
r
k
s
h
o
p
o
p
er
atin
g
in
th
e
a
u
to
m
o
tiv
e
in
d
u
s
tr
y
,
th
e
co
m
p
an
y
is
lo
ca
ted
in
n
o
r
th
o
f
Mo
r
o
cc
o
.
T
h
e
d
ataset
co
n
s
is
ts
o
f
in
f
o
r
m
atio
n
o
n
7
0
u
n
i
q
u
e
m
o
ld
s
,
ea
ch
with
v
ar
y
in
g
s
izes a
n
d
weig
h
ts
.
Ad
d
itio
n
ally
,
it in
clu
d
es f
iv
e
ty
p
e
s
o
f
m
ac
h
in
es with
d
if
f
er
en
t c
l
am
p
in
g
ca
p
ac
ities
,
r
an
g
in
g
f
r
o
m
1
5
0
to
n
s
to
4
0
0
t
o
n
s
.
Fin
ally
,
th
e
d
ataset
co
n
tai
n
s
d
etails
o
n
1
0
d
if
f
er
en
t
t
y
p
es
o
f
p
last
ic
m
ater
ials
.
T
ab
le
1
d
escr
ib
es
th
e
m
ain
u
s
ed
v
ar
iab
les
o
n
th
is
s
tu
d
y
.
T
h
e
d
ata
is
class
if
ied
i
n
to
f
o
u
r
g
r
o
u
p
s
:
p
ar
t_
d
ata,
m
ater
ial_
d
ata,
m
o
ld
_
d
ata,
a
n
d
m
ac
h
in
e_
d
ata.
T
h
e
"p
a
r
t_
d
ata"
co
n
s
o
lid
ates
in
f
o
r
m
atio
n
p
er
tain
in
g
t
o
th
e
ch
ar
ac
ter
is
tics
o
f
th
e
p
ar
t
.
T
h
i
s
in
clu
d
es d
etails lik
e
th
e
p
ar
t
weig
h
t,
s
p
r
u
e
weig
h
t,
an
d
th
e
n
u
m
b
er
o
f
p
ar
ts
p
e
r
ca
v
ity
.
I
t
en
c
o
m
p
ass
es
tr
ee
s
p
ec
if
ic
item
s
.
T
h
e
"m
ater
ial_
d
a
ta"
f
o
cu
s
es
o
n
th
e
ty
p
e
o
f
m
at
er
ial
u
s
ed
,
s
u
ch
as
PA6
,
p
o
ly
p
r
o
p
y
len
e
(
PP
)
,
p
o
l
y
eth
y
len
e
(
PE)
,
a
n
d
o
th
e
r
s
.
I
n
th
e
f
in
al
d
ataset,
th
er
e
ar
e
1
0
d
if
f
er
en
t
m
ater
ial
ty
p
es,
an
d
T
ab
le
1
illu
s
tr
ates
th
r
ee
m
ain
ch
ar
ac
te
r
is
tics
o
f
th
e
m
ater
ials
.
T
h
e
"m
o
ld
d
ata
"
r
eg
r
o
u
p
s
in
f
o
r
m
atio
n
r
elate
d
to
th
e
d
im
en
s
io
n
s
o
f
t
h
e
m
o
ld
,
in
cl
u
d
in
g
its
weig
h
t
,
h
eig
h
t,
len
g
th
,
an
d
wid
th
.
T
h
e
"m
ac
h
in
e_
d
ata"
r
ef
er
s
to
th
e
m
ac
h
in
e
ass
ig
n
e
d
to
ea
ch
p
r
o
d
u
ct
b
ased
o
n
t
h
e
clam
p
in
g
f
o
r
ce
r
e
q
u
ir
ed
,
ch
ar
ac
ter
ized
b
y
th
e
m
ax
im
u
m
f
o
r
ce
n
ec
ess
ar
y
f
o
r
clam
p
in
g
an
d
h
o
ld
in
g
.
T
h
e
u
n
it o
f
m
ea
s
u
r
em
en
t is
k
ilo
-
n
ewt
o
n
o
r
to
n
s
.
T
ab
le
1
.
Data
s
et
item
s
s
p
ec
if
icatio
n
s
D
a
t
a
I
t
e
m
D
e
scri
p
t
i
o
n
P
a
r
t
d
a
t
a
W
e
i
g
h
t
o
f
t
h
e
p
a
r
t
W
e
i
g
h
t
o
f
t
h
e
p
l
a
s
t
i
c
p
a
r
t
[
g
]
S
p
r
u
e
w
e
i
g
h
t
N
u
mb
e
r
o
f
c
a
v
i
t
y
m
o
l
d
[
g
]
To
t
a
l
w
e
i
g
h
t
w
e
i
g
h
t
o
f
t
h
e
p
a
r
t
a
n
d
t
h
e
s
p
r
u
e
[
g
]
M
o
l
d
d
a
t
a
W
i
d
t
h
W
i
d
t
h
d
i
me
n
s
i
o
n
o
f
t
h
e
m
o
l
d
[
mm
]
Le
n
g
t
h
Le
n
g
t
h
o
f
t
h
e
m
o
l
d
[
m
m]
H
e
i
g
h
t
H
e
i
g
h
t
d
i
m
e
n
s
i
o
n
o
f
t
h
e
m
o
l
d
[
mm
]
W
e
i
g
h
t
To
t
a
l
w
e
i
g
h
t
o
f
t
h
e
m
o
l
d
[
K
g
]
M
a
t
e
r
i
a
l
d
a
t
a
d
e
n
si
t
y
D
e
n
si
t
y
o
f
ma
t
e
r
i
a
l
[
K
g
/
m3
]
R
a
t
e
F
l
o
w
F
l
o
w
r
a
t
e
o
f
mat
e
r
i
a
l
[
g
/
1
0
mi
n
]
Ty
p
e
Ty
p
e
o
f
m
a
t
e
r
i
a
l
(
P
P
,
P
E,
a
n
d
P
A
6
)
M
a
c
h
i
n
e
d
a
t
a
S
i
z
e
M
a
c
h
i
n
e
s
i
z
e
,
1
5
0
T
,
2
0
0
T
,
2
5
0
T
,
3
2
0
T,
4
0
0
T
3.
3
.
Da
t
a
a
na
ly
s
is
T
h
e
i
n
i
t
i
a
l
s
t
e
p
o
f
t
h
e
i
n
v
e
s
t
i
g
a
ti
o
n
i
n
v
o
l
v
e
s
c
o
n
d
u
c
t
i
n
g
a
c
o
r
r
e
l
a
t
i
o
n
s
t
u
d
y
b
e
t
w
e
e
n
t
h
e
c
l
as
s
es
a
n
d
t
h
e
i
r
r
e
s
p
e
c
t
i
v
e
f
e
a
t
u
r
es
.
C
o
r
r
e
la
t
i
o
n
,
b
r
o
a
d
l
y
d
e
f
i
n
e
d
,
q
u
a
n
t
i
f
i
e
s
t
h
e
r
e
l
a
t
i
o
n
s
h
i
p
b
e
t
w
ee
n
v
a
r
i
a
b
l
es
.
I
n
c
o
r
r
e
l
a
t
e
d
d
a
ta
,
a
l
t
e
r
a
ti
o
n
s
i
n
t
h
e
m
a
g
n
i
t
u
d
e
o
f
o
n
e
v
a
r
i
a
b
l
e
a
r
e
l
i
n
k
e
d
t
o
c
h
a
n
g
e
s
i
n
t
h
e
m
a
g
n
i
t
u
d
e
o
f
a
n
o
t
h
er
v
a
r
i
a
b
l
e
,
e
i
t
h
e
r
i
n
t
h
e
s
a
m
e
d
i
r
ec
t
i
o
n
(
p
o
s
it
i
v
e
c
o
r
r
e
l
a
t
i
o
n
)
o
r
i
n
t
h
e
o
p
p
o
s
it
e
d
i
r
e
c
t
i
o
n
(
n
e
g
a
t
i
v
e
c
o
r
r
e
la
t
i
o
n
)
[
1
9
]
.
T
h
e
co
r
r
elatio
n
m
atr
ix
c
o
n
f
ir
m
s
a
s
ig
n
if
ican
t r
elatio
n
b
etwe
en
clam
p
in
g
f
o
r
ce
an
d
f
ea
tu
r
es
r
elate
d
t
o
m
o
ld
d
im
en
s
io
n
s
,
with
s
tr
o
n
g
p
o
s
itiv
e
c
o
ef
f
icien
ts
class
if
ied
as
f
o
llo
ws:
m
o
ld
weig
h
t
(
0
.
7
8
8
)
,
m
o
ld
h
eig
h
t
(
0
.
7
3
7
)
,
m
o
l
d
len
g
th
(
0
.
7
4
1
)
,
an
d
m
o
ld
wid
th
(
0
.
7
6
1
)
.
I
n
d
ee
d
,
th
e
s
elec
tio
n
o
f
m
ac
h
in
e
s
ize
b
ased
o
n
th
e
p
ar
t
is
s
ig
n
if
ican
tly
in
f
lu
en
ce
d
b
y
th
e
d
im
en
s
io
n
s
an
d
ch
ar
ac
te
r
is
tics
o
f
th
e
m
o
ld
as
d
i
s
p
lay
ed
o
n
th
e
F
ig
u
r
e
4
s
h
o
win
g
th
e
m
ac
h
in
e
allo
ca
ti
o
n
ac
co
r
d
in
g
to
th
e
s
ize
o
f
t
h
e
m
o
ld
.
I
t
is
h
ig
h
l
y
lik
ely
t
h
at
s
m
all
m
o
ld
s
ar
e
ad
d
r
ess
ed
to
s
m
aller
m
ac
h
i
n
e
s
izes,
wh
ile
lar
g
e
m
o
ld
s
ar
e
al
lo
ca
ted
to
lar
g
er
m
ac
h
i
n
e
s
ize
s
.
Fu
r
th
er
m
o
r
e,
t
h
e
co
r
r
elatio
n
co
ef
f
icie
n
t in
d
icate
s
th
at
clam
p
in
g
f
o
r
ce
is
in
f
lu
e
n
ce
d
b
y
f
ea
tu
r
es a
s
s
o
ciate
d
als
o
with
p
ar
t w
eig
h
t,
with
a
p
o
s
i
tiv
e
co
ef
f
icien
t
o
f
(
0
.
5
9
1
)
f
o
r
to
tal
p
ar
t
weig
h
t
an
d
(
0
.
4
)
f
o
r
p
ar
t
weig
h
t.
On
th
e
o
th
er
h
an
d
,
f
ea
tu
r
es
r
elate
d
to
p
last
ic
ty
p
es,
s
u
c
h
as
d
en
s
ity
,
m
elt
f
lo
w,
an
d
m
ater
ial
ty
p
e
,
d
e
m
o
n
s
tr
ate
lo
wer
co
r
r
elatio
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Ma
ch
in
e
lea
r
n
in
g
-
a
s
s
is
ted
d
ec
is
io
n
s
u
p
p
o
r
t in
in
d
u
s
tr
ia
l .
.
.
(
F
a
o
u
z
i Ta
ya
la
ti)
275
Sp
ec
if
ically
,
th
e
co
r
r
elatio
n
c
o
ef
f
icien
ts
f
o
r
th
ese
f
ea
tu
r
es
a
r
e
(
0
.
1
5
6
)
f
o
r
d
e
n
s
ity
,
(
-
0
.
3
4
6
)
f
o
r
m
elt
f
lo
w,
an
d
(
0
.
1
1
)
f
o
r
m
ater
ial
ty
p
e.
T
h
is
in
itial
an
aly
s
is
co
n
f
ir
m
s
th
at
m
o
ld
a
n
d
p
ar
t
f
ea
tu
r
es
ar
e
t
h
e
m
o
s
t
s
ig
n
if
ican
t
co
n
tr
ib
u
to
r
s
to
m
ac
h
in
e
s
elec
ti
o
n
.
T
h
is
is
co
n
s
is
ten
t
with
th
e
ac
tu
al
p
last
ic
m
o
ld
i
n
g
m
an
u
f
ac
tu
r
in
g
.
T
h
e
v
is
u
aliza
tio
n
in
Fig
u
r
e
5
d
e
m
o
n
s
tr
ates
th
e
i
n
ter
ac
tio
n
a
m
o
n
g
th
e
v
ar
iab
les
o
f
th
e
d
ataset,
s
p
ec
if
ically
r
ep
r
esen
ted
b
y
th
e
p
ar
am
eter
"
m
ac
h
in
e
clam
p
in
g
f
o
r
ce
,
"
in
r
elatio
n
th
e
m
o
ld
s
p
ec
f
icatio
n
a
n
d
p
ar
t w
eig
h
t
.
Fig
u
r
e
4
.
Data
co
r
r
elatio
n
m
at
r
ix
(
m
ac
h
i
n
e
clam
p
in
g
f
o
r
ce
v
s
d
ataset
f
ea
tu
r
es)
Fig
u
r
e
5
.
C
lass
if
icatio
n
o
f
th
e
m
ac
h
in
e
s
ize
ac
co
r
d
i
n
g
to
t
h
e
m
o
ld
s
p
ec
f
icatio
n
a
n
d
p
ar
t w
e
ig
h
t
T
h
e
d
is
tr
ib
u
tio
n
o
f
class
es
b
ased
o
n
th
e
m
ain
f
ea
tu
r
es
m
o
ld
an
d
p
ar
t
weig
h
t
is
illu
s
tr
ated
i
n
F
ig
u
r
e
5
.
T
h
e
d
is
tr
ib
u
tio
n
s
ar
e
d
is
tin
ctl
y
r
ep
r
esen
ted
b
y
f
i
v
e
class
es:
C
las
s
1
(
1
5
0
t)
,
C
lass
2
(
2
0
0
t)
,
C
lass
3
(
2
5
0
t)
,
C
las
s
4
(
3
2
0
t)
,
an
d
C
lass
5
(
4
0
0
t)
.
T
h
e
m
ac
h
in
e
d
is
tr
ib
u
tio
n
f
o
llo
ws
a
n
o
r
m
al
d
is
tr
ib
u
tio
n
,
with
th
e
av
er
ag
e
ce
n
ter
ed
o
n
m
ac
h
in
es
2
0
0
t
a
n
d
2
5
0
t.
T
h
e
tr
en
d
o
b
s
er
v
ed
in
d
icate
s
th
at
s
m
aller
m
o
l
d
s
a
r
e
m
o
r
e
lik
ely
to
b
e
ass
ig
n
ed
to
s
m
aller
m
ac
h
in
es.
Fo
r
in
s
tan
ce
,
m
o
ld
s
with
a
h
eig
h
t
less
th
an
4
0
0
m
m
ten
d
to
b
e
ass
ig
n
ed
to
C
las
s
1
an
d
2
m
ac
h
in
es,
wh
ile
m
o
ld
s
with
a
h
eig
h
t
ex
ce
ed
in
g
7
0
0
m
m
ar
e
ass
o
ciate
d
with
C
las
s
5
m
ac
h
in
es.
T
h
is
co
r
r
elatio
n
u
n
d
er
s
co
r
es
th
e
im
p
ac
t
o
f
m
o
ld
d
im
e
n
s
io
n
s
o
n
th
e
allo
ca
tio
n
o
f
m
ac
h
in
es.
Mo
r
eo
v
er
,
th
e
p
ar
am
eter
"M
o
l
d
W
id
th
"
s
ig
n
if
ican
tly
in
f
lu
e
n
ce
s
th
e
s
elec
tio
n
o
f
th
e
m
ac
h
in
e.
Sp
ec
if
icall
y
,
m
ac
h
in
es
with
a
wid
th
less
th
an
4
0
0
m
m
ar
e
ass
ig
n
ed
to
s
m
aller
m
ac
h
in
e
class
es (
C
las
s
1
an
d
2
)
,
wh
ile
m
a
ch
in
es with
a
wid
th
ex
ce
ed
in
g
6
0
0
m
m
ar
e
allo
ca
t
ed
to
m
ac
h
in
es
o
f
2
5
0
t,
3
2
0
t,
an
d
4
0
0
t.
T
h
is
p
ar
am
eter
is
d
ir
ec
tly
lin
k
ed
to
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
270
-
2
8
5
276
tie
b
ar
o
f
th
e
m
ac
h
in
e
a
n
d
s
er
v
es a
s
a
cr
itical
ch
a
r
ac
ter
is
tic
d
eter
m
in
in
g
wh
eth
er
th
e
m
o
ld
ca
n
f
it th
e
m
ac
h
in
e
.
L
o
g
ically
,
if
th
e
m
o
ld
h
as
s
u
b
s
tan
tial
wid
th
d
im
en
s
io
n
s
,
it
is
as
s
ig
n
ed
to
a
m
ac
h
in
e
with
a
h
ig
h
er
d
is
tan
ce
b
etwe
en
tie
b
ar
s
.
C
o
n
ce
r
n
in
g
t
h
e
to
tal
weig
h
t
o
f
th
e
p
ar
t,
th
is
f
ea
tu
r
e
ex
h
ib
its
a
s
ig
n
if
ican
t
i
m
p
ac
t
o
n
t
h
e
ch
o
ice
o
f
th
e
m
ac
h
in
e.
Par
ts
with
a
weig
h
t
ex
ce
ed
in
g
3
0
0
g
ar
e
ty
p
ically
ass
ig
n
ed
to
Ma
ch
in
e
5
,
wh
ile
p
ar
ts
with
weig
h
ts
r
an
g
in
g
b
etwe
en
2
0
0
g
an
d
3
0
0
g
ar
e
allo
ca
te
d
to
m
ac
h
in
e
2
an
d
m
ac
h
in
e
4
.
T
h
is
in
d
icate
s
a
c
lear
co
r
r
elatio
n
b
etwe
en
th
e
t
o
tal
p
last
ic
weig
h
t
an
d
th
e
s
elec
tio
n
o
f
s
p
ec
if
ic
m
ac
h
i
n
es,
h
ig
h
li
g
h
t
in
g
th
e
im
p
o
r
tan
ce
o
f
th
is
f
ea
tu
r
e
in
th
e
d
ec
is
io
n
-
m
ak
in
g
p
r
o
ce
s
s
.
T
h
e
in
itial
cr
o
s
s
-
ch
ec
k
o
f
th
e
d
ataset
r
ev
ea
ls
co
m
p
ellin
g
ev
id
en
ce
in
d
icatin
g
th
at
f
ea
tu
r
es
ass
o
ciate
d
with
b
o
th
th
e
s
i
ze
o
f
th
e
m
o
ld
an
d
t
h
e
to
tal
w
eig
h
t
o
f
t
h
e
p
last
ic
p
ar
t e
x
er
t a
s
ig
n
i
f
ican
t im
p
ac
t
o
n
th
e
d
ec
is
io
n
-
m
ak
in
g
p
r
o
ce
s
s
r
eg
ar
d
in
g
m
ac
h
in
e
allo
c
atio
n
.
T
h
e
d
ata
u
s
ed
in
th
is
s
tu
d
y
ex
h
ib
its
n
o
n
-
e
q
u
itab
le
d
is
tr
ib
u
tio
n
am
o
n
g
m
ac
h
i
n
es,
as
d
ep
icted
in
F
ig
u
r
e
6
.
No
tab
ly
i
n
Fig
u
r
e
6
(
a)
,
m
ac
h
i
n
e
2
0
0
t
(
C
lass
2
)
h
as
th
e
h
ig
h
est
r
ep
r
esen
tatio
n
with
2
0
r
ec
o
r
d
s
,
f
o
llo
wed
b
y
Ma
c
h
in
e
3
with
1
4
r
ec
o
r
d
s
,
a
n
d
th
e
r
em
ai
n
in
g
m
ac
h
in
es
ea
ch
h
av
in
g
1
2
r
ec
o
r
d
s
.
Ma
ch
in
e
2
0
0
t
s
tan
d
s
o
u
t
as
th
e
m
ajo
r
ity
class
,
wh
ile
Ma
ch
in
es
4
,
5
,
an
d
1
ar
e
co
n
s
id
er
ed
m
in
o
r
ity
class
es.
T
o
ad
d
r
ess
th
is
im
b
alan
ce
,
th
e
f
ir
s
t
s
tep
in
v
o
lv
ed
eq
u
ilib
r
atin
g
th
e
d
ataset
u
s
in
g
th
e
o
v
er
s
am
p
lin
g
tech
n
iq
u
e
[
2
0
]
.
Ad
d
itio
n
al
r
ec
o
r
d
s
wer
e
g
en
e
r
ated
f
o
r
th
e
m
in
o
r
ity
m
ac
h
in
es,
en
s
u
r
in
g
th
at
ea
ch
m
ac
h
in
e
n
o
w
h
as
2
0
r
ec
o
r
d
s
.
C
o
n
s
eq
u
en
tly
,
th
e
f
in
al
d
atase
t h
as b
ee
n
ex
p
an
d
e
d
to
[
1
0
0
×
1
1
]
as sh
o
wn
in
Fig
u
r
e
6
(
b
)
.
(
a)
(
b
)
Fig
u
r
e
6
.
Data
b
ala
n
cin
g
:
(
a
)
i
m
b
alan
cin
g
d
ataset
an
d
(
b
)
b
al
an
ce
d
d
ataset
3
.
4
.
M
a
chine le
a
rning
m
et
h
o
ds
a
nd
m
ulticr
it
er
ia
pro
blem
T
h
e
p
r
o
b
lem
o
f
m
ac
h
i
n
e
s
elec
tio
n
is
co
n
s
id
er
ed
as
a
m
u
lti
-
cr
iter
ia
p
r
o
b
lem
,
with
its
m
u
ltip
le
cr
iter
ia
an
d
p
ar
a
m
eter
s
,
alig
n
s
well
with
th
e
ca
p
ab
ilit
ies
o
f
class
if
icatio
n
alg
o
r
ith
m
s
.
T
h
ese
alg
o
r
it
h
m
s
ar
e
d
esig
n
ed
t
o
h
an
d
le
m
u
lti
-
cr
iter
ia
p
r
o
b
lem
s
b
y
lear
n
i
n
g
f
r
o
m
d
iv
e
r
s
e
s
ets
o
f
in
p
u
t
f
ea
tu
r
es
an
d
m
a
k
in
g
p
r
e
d
ictio
n
s
o
r
class
if
icatio
n
s
b
ased
o
n
th
e
le
ar
n
ed
p
atter
n
s
.
I
n
t
h
e
co
n
te
x
t
o
f
th
is
s
tu
d
y
,
t
h
e
class
if
icati
o
n
alg
o
r
ith
m
s
wer
e
ap
p
lied
to
p
r
ed
ict
th
e
"c
lam
p
in
g
f
o
r
c
e
m
ac
h
in
e"
b
y
c
o
n
s
id
er
in
g
v
ar
io
u
s
p
ar
a
m
ete
r
s
r
elate
d
to
m
ac
h
in
e,
m
ater
ial,
m
o
ld
,
an
d
p
ar
t
ch
ar
ac
ter
is
tics
.
T
h
is
ap
p
r
o
ac
h
all
o
ws
f
o
r
a
s
y
s
tem
atic
an
d
d
ata
-
d
r
iv
e
n
d
ec
is
io
n
-
m
a
k
in
g
p
r
o
ce
s
s
in
th
e
co
m
p
lex
d
o
m
ai
n
o
f
m
ac
h
in
e
s
elec
tio
n
.
Six
al
g
o
r
ith
m
s
wer
e
ap
p
lied
:
RF
,
DT
,
SVM
,
NB
,
SG
D
c
lass
if
ier
,
an
d
LR
.
T
h
e
g
o
al
is
to
an
ticip
ate
th
e
"c
lam
p
in
g
f
o
r
ce
m
ac
h
in
e
.
3
.
4
.
1
.
Ra
nd
o
m
f
o
re
s
t
R
F
i
s
co
n
s
id
er
ed
a
cr
u
cial
ML
alg
o
r
ith
m
f
o
r
p
atter
n
r
e
co
g
n
itio
n
,
p
r
im
ar
ily
b
ec
au
s
e
o
f
its
co
s
t
-
ef
f
ec
tiv
en
ess
.
T
h
e
f
u
n
d
am
en
t
al
p
r
in
cip
le
b
eh
in
d
its
tr
ain
in
g
s
tr
ateg
y
is
b
a
g
g
in
g
[
2
1
]
.
T
h
is
s
u
g
g
ests
th
at
RF
is
b
u
ilt
u
p
o
n
en
s
em
b
le
s
am
p
lin
g
with
o
u
t
r
ep
lace
m
e
n
t
f
r
o
m
s
u
b
s
ets
o
f
th
e
d
ataset.
T
h
e
r
e
m
a
in
in
g
d
ata,
ter
m
ed
out
-
of
-
b
ag
,
a
r
e
u
s
ed
f
o
r
ass
ess
in
g
th
e
m
o
d
el'
s
p
er
f
o
r
m
an
ce
.
Ma
n
y
b
o
o
s
tin
g
o
r
b
ag
g
in
g
alg
o
r
ith
m
s
ar
e
ce
n
ter
e
d
ar
o
u
n
d
DT
.
T
h
e
in
itial
n
o
d
e'
s
s
tate
g
en
er
ates
ad
d
itio
n
al
n
o
d
es
co
n
tain
in
g
f
ea
tu
r
es
d
ir
ec
ted
u
p
war
d
.
As
a
r
esu
lt,
m
u
ltip
le
DT
ar
e
em
p
lo
y
ed
to
c
lass
if
y
ea
ch
d
ataset
with
s
am
p
l
in
g
th
is
m
eth
o
d
,
in
d
iv
i
d
u
al
DT
m
ay
e
x
h
ib
it
l
o
wer
ac
cu
r
ac
y
co
m
p
ar
ed
to
a
DT
c
o
n
s
tr
u
cted
u
s
in
g
th
e
e
n
tire
d
ataset.
T
h
er
ef
o
r
e,
it
is
ad
v
an
tag
e
o
u
s
to
ag
g
r
e
g
ate
th
e
r
esu
lts
o
f
ea
ch
tr
ee
,
as th
e
y
ca
n
co
llectiv
ely
class
if
y
tr
ain
ed
d
ata
in
a
co
m
p
lem
en
tar
y
m
an
n
er
.
3
.
4
.
2
.
Dec
is
io
n t
re
e
T
h
is
m
o
d
el
class
if
ies d
ata
o
r
p
r
ed
icts
o
u
tco
m
es b
y
s
p
litt
in
g
a
d
ataset
b
ased
o
n
r
esp
o
n
s
es to
q
u
esti
o
n
s
[
2
2
]
.
As
a
s
u
p
er
v
is
ed
lear
n
in
g
m
eth
o
d
,
it
is
tr
ain
ed
o
n
a
d
atas
et
with
k
n
o
wn
tar
g
et
class
if
icatio
n
s
.
T
h
is
tr
ain
in
g
allo
ws th
e
m
o
d
el
to
m
a
k
e
ac
c
u
r
ate
p
r
e
d
ictio
n
s
o
n
n
ew
d
ata.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Ma
ch
in
e
lea
r
n
in
g
-
a
s
s
is
ted
d
ec
is
io
n
s
u
p
p
o
r
t in
in
d
u
s
tr
ia
l .
.
.
(
F
a
o
u
z
i Ta
ya
la
ti)
277
3
.
4
.
3
.
Na
iv
es
B
a
y
es m
o
del
T
h
e
n
aïv
e
B
ay
es
alg
o
r
ith
m
is
a
s
u
p
er
v
is
ed
lear
n
in
g
m
eth
o
d
b
ased
o
n
B
ay
es'
th
eo
r
em
,
u
s
ed
f
o
r
class
if
icatio
n
p
r
o
b
lem
s
[
2
3
]
.
I
t
ac
ts
as
a
p
r
o
b
ab
ilis
tic
class
if
ie
r
,
p
r
e
d
ictin
g
th
e
lik
elih
o
o
d
o
f
a
n
o
b
ject
b
elo
n
g
in
g
to
a
s
p
ec
if
ic
class
.
T
h
is
m
ak
es it e
f
f
icien
t f
o
r
h
an
d
lin
g
lar
g
e
d
atasets
.
3
.
4
.
4
.
Sto
cha
s
t
ic
g
ra
dient
de
s
ce
nt
cla
s
s
if
ier
SGD
i
s
a
p
o
wer
f
u
l
alg
o
r
ith
m
u
s
ed
to
f
in
d
p
ar
am
eter
v
alu
es
th
at
m
in
im
ize
a
co
s
t
f
u
n
ctio
n
[
2
4
]
.
I
t
is
wid
ely
ap
p
lied
in
d
is
cr
im
in
ativ
e
lear
n
in
g
task
s
in
v
o
lv
in
g
l
in
ea
r
class
if
ier
s
.
I
t
s
ef
f
ec
tiv
en
ess
is
p
ar
ticu
lar
ly
n
o
tab
le
with
co
n
v
ex
l
o
s
s
f
u
n
ct
io
n
s
s
u
ch
as SVM
an
d
LR
.
3
.
4
.
5
.
L
o
g
is
t
ic
re
g
re
s
s
io
n
T
h
e
m
o
ld
el
is
u
s
ed
f
o
r
f
o
r
ec
as
tin
g
th
e
v
alu
es
o
f
a
s
p
ec
if
ic
v
a
r
iab
le
b
y
c
o
n
s
id
er
in
g
th
e
v
al
u
es o
f
o
th
er
v
ar
iab
les.
An
y
ch
a
n
g
e
i
n
p
r
ed
ictio
n
is
ter
m
ed
as
th
e
d
ep
e
n
d
en
t
v
ar
iab
le
[
2
5
]
.
An
in
d
ep
en
d
en
t
v
a
r
iab
le
is
o
n
e
th
at
is
u
s
ed
to
p
r
ed
ict
th
e
v
alu
es
o
f
o
th
er
v
ar
ia
b
les.
Fo
r
in
s
tan
ce
,
in
th
e
co
n
te
x
t
o
f
lin
ea
r
r
e
g
r
ess
io
n
,
we
u
tili
ze
in
d
ep
en
d
en
t v
ar
ia
b
les to
u
n
d
e
r
s
tan
d
h
o
w
ch
a
n
g
es in
th
e
m
r
e
late
to
ch
an
g
es in
th
e
d
ep
e
n
d
e
n
t v
ar
iab
le.
3
.
4
.
6
.
Su
pp
o
rt
v
ec
t
o
r
m
a
chi
ne
SVM
p
o
s
s
es
s
v
er
s
atility
in
h
an
d
lin
g
b
o
th
r
e
g
r
ess
io
n
an
d
class
if
icatio
n
task
s
.
Ho
we
v
er
,
th
eir
p
er
f
o
r
m
an
ce
ten
d
s
to
ex
ce
l
p
a
r
ticu
lar
ly
in
class
if
icatio
n
p
r
o
b
lem
s
.
SVM
g
ain
ed
wid
esp
r
ea
d
p
o
p
u
lar
ity
u
p
o
n
th
eir
cr
ea
tio
n
an
d
r
em
ain
a
p
r
o
m
in
en
t
ch
o
ice
f
o
r
h
i
g
h
-
p
er
f
o
r
m
in
g
alg
o
r
ith
m
s
,
o
f
ten
r
eq
u
ir
in
g
m
in
im
al
tu
n
in
g
to
ac
h
iev
e
ex
ce
llen
t
r
esu
lts
.
I
t
'
s
a
s
u
p
er
v
is
ed
ML
ap
p
r
o
ac
h
wh
er
e
th
e
o
b
jectiv
e
is
to
d
ete
r
m
in
e
a
h
y
p
e
r
p
lan
e
th
at
ef
f
ec
tiv
ely
s
eg
r
eg
ates
two
class
es
in
th
e
d
ataset.
B
o
th
S
VM
an
d
L
R
aim
to
id
e
n
tify
t
h
e
o
p
ti
m
al
h
y
p
er
p
lan
e
f
o
r
class
if
icatio
n
task
s
.
Ho
wev
er
,
L
R
is
a
p
r
o
b
ab
ilis
tic
ap
p
r
o
a
ch
,
esti
m
atin
g
th
e
p
r
o
b
ab
ilit
y
t
h
at
a
g
iv
e
n
in
s
tan
ce
b
elo
n
g
s
to
a
p
ar
ticu
lar
class
,
wh
ile
SVM
is
r
o
o
ted
i
n
s
tatis
ti
ca
l p
r
in
cip
les,
f
o
cu
s
in
g
o
n
m
a
x
im
izin
g
th
e
m
ar
g
in
b
etwe
en
class
es f
o
r
ef
f
ec
tiv
e
s
ep
ar
atio
n
[
2
6
]
.
3
.
4
.
7
.
P
er
f
o
r
m
a
nce
m
et
rics a
nd
co
nfusi
o
n m
a
t
rix
T
h
e
tech
n
iq
u
e
em
p
l
o
y
ed
f
o
r
ass
es
s
in
g
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
s
ix
m
ac
h
in
es
in
v
o
lv
es
th
e
u
s
e
o
f
a
co
n
f
u
s
io
n
m
atr
ix
[
2
7
]
.
T
h
is
m
atr
ix
p
r
o
v
id
es
a
ta
b
u
lar
r
ep
r
esen
tatio
n
o
f
v
ar
io
u
s
o
u
tco
m
e
s
in
a
class
if
icatio
n
p
r
o
b
lem
,
aid
i
n
g
in
th
e
v
is
u
aliza
tio
n
o
f
p
r
ed
ictio
n
r
esu
lts
.
I
t
o
u
tlin
es
a
tab
le
wi
th
th
e
p
r
ed
icted
an
d
ac
tu
al
v
alu
es
o
f
a
class
if
ier
,
f
o
llo
win
g
th
e
s
tan
d
ar
d
r
ep
r
esen
tat
io
n
illu
s
tr
ated
in
T
ab
le
2
.
−
T
r
u
e
p
o
s
itiv
e
:
th
is
is
also
k
n
o
wn
as
th
e
co
u
n
t
o
f
t
r
u
e
p
o
s
itiv
es,
wh
ich
s
ig
n
if
ies
th
e
n
u
m
b
er
o
f
in
s
tan
ce
s
wh
er
e
th
e
m
o
d
el'
s
p
r
ed
ictio
n
o
f
a
p
o
s
itiv
e
o
u
tco
m
e
alig
n
s
with
th
e
ac
tu
al
p
o
s
itiv
e
o
u
tco
m
e
.
−
Fals
e
p
o
s
itiv
e
:
th
is
is
ter
m
ed
a
s
f
alse
p
o
s
itiv
e
in
b
in
a
r
y
class
if
icatio
n
.
I
t
r
ep
r
esen
ts
th
e
in
s
t
an
ce
s
wh
er
e
t
h
e
m
o
d
el
in
co
r
r
ec
tly
p
r
ed
icts
th
e
p
o
s
itiv
e
class
wh
en
th
e
ac
tu
al
class
is
n
eg
ativ
e.
−
T
r
u
e
n
e
g
ativ
e
:
it
r
ep
r
esen
ts
th
e
in
s
tan
ce
s
wh
er
e
th
e
m
o
d
el
c
o
r
r
ec
tly
p
r
ed
icts
th
e
n
eg
ativ
e
class
wh
en
th
e
ac
tu
al
class
is
al
s
o
n
eg
ativ
e
.
−
Fals
e
n
eg
ativ
e
:
it
s
ig
n
if
ies
th
e
in
s
tan
ce
s
wh
er
e
th
e
m
o
d
el
in
co
r
r
ec
tly
p
r
ed
icts
th
e
n
eg
ativ
e
class
wh
en
th
e
ac
tu
al
class
is
p
o
s
itiv
e.
a)
Acc
u
r
ac
y
:
t
h
e
ac
cu
r
ac
y
,
wh
ic
h
m
ea
s
u
r
es
th
e
p
r
o
p
o
r
tio
n
o
f
c
o
r
r
ec
t
p
r
ed
ictio
n
s
o
u
t
o
f
all
p
r
ed
ictio
n
s
m
ad
e
b
y
th
e
m
o
d
el.
W
e
ca
n
ca
lcu
lat
e
th
e
m
ea
n
a
b
s
o
lu
te
er
r
o
r
(
M
AE
)
u
s
in
g
th
e
(
1
)
.
=
+
+
+
+
(
1
)
b)
Pre
cisi
o
n
:
Pre
cisi
o
n
is
a
m
ea
s
u
r
e
o
f
th
e
ac
cu
r
ac
y
o
f
th
e
p
o
s
iti
v
e
p
r
ed
ictio
n
s
m
ad
e
b
y
a
m
o
d
e
l.
T
h
e
f
o
r
m
u
l
a
is
(
2
)
.
=
+
(
2
)
c)
R
ec
all/Sen
s
itiv
ity
:
R
ec
a
ll,
als
o
k
n
o
wn
as
s
en
s
itiv
ity
,
q
u
an
tif
ies th
e
p
r
o
p
o
r
tio
n
o
f
p
o
s
itiv
e
ca
s
es c
o
r
r
ec
tly
id
en
tifie
d
b
y
th
e
class
if
ier
am
o
n
g
all
ac
tu
al
p
o
s
itiv
e
ca
s
es in
th
e
d
ataset.
I
ts
f
o
r
m
u
la
is
ex
p
r
ess
ed
as
(
3
)
.
=
+
(
3
)
d)
F1
-
s
co
r
e
:
T
h
e
F1
-
s
co
r
e,
a
m
e
asu
r
e
th
at
co
m
b
in
es
p
r
ec
is
io
n
an
d
r
ec
all,
is
o
f
ten
d
ef
in
ed
as
th
e
h
ar
m
o
n
ic
m
ea
n
o
f
th
e
two
m
etr
ics.
Har
m
o
n
ic
m
ea
n
is
f
av
o
r
e
d
f
o
r
r
at
io
s
,
lik
e
p
r
ec
is
io
n
an
d
r
ec
all,
as
it
b
alan
ce
s
ex
tr
em
e
v
al
u
es
m
o
r
e
ef
f
ec
tiv
ely
th
an
th
e
ar
ith
m
etic
m
ea
n
.
T
h
e
f
o
r
m
u
la
f
o
r
ca
lcu
latin
g
t
h
e
F1
-
s
co
r
e
is
g
iv
en
b
y
(
4
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
270
-
2
8
5
278
1
−
=
2
∗
∗
∗
(
4
)
T
ab
le
2
.
C
o
n
f
u
s
io
n
m
atr
ix
r
ep
r
esen
tatio
n
A
c
t
u
a
l
P
o
si
t
i
v
e
N
e
g
a
t
i
v
e
P
r
e
d
i
c
t
e
d
P
o
si
t
i
v
e
Tr
u
e
p
o
si
t
i
v
e
F
a
l
se
p
o
si
t
i
v
e
N
e
g
a
t
i
v
e
F
a
l
se
n
e
g
a
t
i
v
e
Tr
u
e
n
e
g
a
t
i
v
e
4.
M
AIN R
E
S
UL
T
S
T
h
e
s
ix
m
o
d
els
u
n
d
e
r
wen
t
tr
a
in
in
g
u
s
in
g
8
5
%
o
f
th
e
d
ataset
an
d
wer
e
s
u
b
s
eq
u
en
tly
test
ed
with
th
e
r
em
ain
in
g
1
5
%.
T
h
is
s
ig
n
if
ies
th
at
th
e
m
o
d
els
wer
e
ex
p
o
s
ed
to
8
5
m
o
ld
s
an
d
p
ar
ts
f
o
r
tr
ai
n
i
n
g
a
n
d
later
ap
p
lied
to
p
r
ed
ict
o
u
tco
m
es
f
o
r
1
5
ad
d
itio
n
al
p
ar
ts
.
T
h
e
r
esu
lts
o
f
th
ese
p
r
ed
ictio
n
s
ar
e
d
etail
ed
in
T
ab
le
3
.
T
h
e
m
eth
o
d
o
l
o
g
y
f
o
llo
wed
to
a
n
al
y
ze
th
e
o
b
tain
e
d
r
esu
lt
is
s
tr
u
ctu
r
ed
in
to
t
h
r
ee
s
ec
tio
n
s
.
Firstl
y
,
a
co
m
p
r
eh
e
n
s
iv
e
tab
le
is
em
p
lo
y
ed
to
ag
g
r
e
g
ate
m
etr
ics
s
u
ch
as
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e.
T
h
i
s
f
ir
s
t
r
esu
lt
s
er
v
es
as
a
co
n
s
o
lid
ated
s
u
m
m
ar
y
o
f
th
e
p
er
f
o
r
m
a
n
ce
in
d
icato
r
s
o
f
th
e
s
ix
m
o
d
els.
Seco
n
d
ly
,
a
co
n
f
u
s
io
n
m
atr
i
x
is
ex
p
lo
r
ed
t
o
ex
am
in
e
th
e
alig
n
m
en
t
b
etwe
en
th
e
ac
tu
al
an
d
p
r
ed
icted
v
alu
es,
o
f
f
er
in
g
i
n
s
ig
h
ts
in
to
th
e
m
o
d
el'
s
class
if
i
ca
tio
n
o
u
tco
m
es.
Fin
all
y
,
th
e
ev
alu
atio
n
in
clu
d
es
th
e
ex
am
in
atio
n
o
f
lear
n
in
g
cu
r
v
e
s
[
2
8
]
to
ass
ess
th
e
tr
ain
in
g
ap
p
r
o
ac
h
es a
d
o
p
ted
b
y
ea
ch
m
o
d
el
o
v
er
th
e
c
o
u
r
s
e
o
f
th
eir
tr
ain
i
n
g
.
T
ab
le
3
.
Per
f
o
r
m
an
ce
an
aly
s
is
M
o
d
e
l
A
c
c
u
r
a
c
y
(
%)
P
r
e
c
i
s
i
o
n
(
%)
R
e
c
a
l
l
(
%)
F1
-
S
c
o
r
e
(
%)
RF
93
96
95
95
DT
87
90
90
90
S
V
M
87
90
90
90
NB
73
67
63
64
LR
53
42
50
45
S
G
D
c
l
a
ss
i
f
i
e
r
27
12
40
18
4
.
1
.
M
et
ric
re
po
rt
(
prec
is
io
n,
re
ca
ll,
F1
,
a
nd
a
cc
ura
c
y
)
I
n
ter
m
s
o
f
ac
cu
r
ac
y
,
th
e
RF
m
o
d
el
ex
h
ib
its
th
e
h
ig
h
est
p
er
f
o
r
m
an
ce
,
ac
h
iev
in
g
a
9
3
%
ac
cu
r
ac
y
r
ate
in
tr
u
e
p
r
ed
ictio
n
s
.
Fo
llo
win
g
clo
s
ely
ar
e
th
e
DT
an
d
SVM
m
o
d
els,
b
o
th
d
em
o
n
s
tr
atin
g
a
co
m
p
ar
ab
le
ac
cu
r
ac
y
o
f
8
7
%.
T
h
e
NB
m
o
d
el
s
h
o
w
s
an
ac
cu
r
ac
y
ar
o
u
n
d
7
3
%,
w
h
ile
th
e
r
em
ai
n
in
g
m
o
d
els,
L
R
an
d
SGD,
d
is
p
lay
lo
wer
ac
cu
r
ac
y
r
ates
at
5
3
%
an
d
2
7
%,
r
esp
ec
tiv
ely
.
I
n
t
er
m
s
o
f
p
r
ec
is
io
n
,
th
e
RF
alg
o
r
ith
m
o
n
ce
a
g
ain
d
em
o
n
s
tr
ates
th
e
b
est
r
esu
lt
with
a
weig
h
ted
av
er
ag
e
o
f
9
5
%
.
T
h
is
in
d
icate
s
th
at
th
e
RF
m
o
d
el
h
as
o
n
ly
a
5
%
er
r
o
r
r
ate
i
n
co
r
r
ec
tly
s
elec
tin
g
th
e
r
ig
h
t
m
ac
h
in
e.
T
h
e
DT
a
n
d
SVM
m
o
d
els
ex
h
ib
it
s
im
ilar
p
r
ec
is
io
n
r
esu
lts
,
b
o
th
ac
h
iev
in
g
a
weig
h
ted
av
e
r
ag
e
o
f
8
7
%.
T
h
e
NB
m
o
d
el
s
h
o
ws
a
p
r
ec
is
io
n
o
f
7
3
%.
H
o
wev
er
,
th
e
p
r
ec
is
io
n
f
o
r
t
h
e
r
em
ai
n
in
g
m
o
d
els,
s
p
ec
if
ically
th
e
SGD
class
if
ier
an
d
LR
,
is
n
o
tab
ly
l
o
wer
,
with
v
alu
es
b
elo
w
4
4
%,
in
d
icatin
g
p
o
o
r
er
p
e
r
f
o
r
m
an
c
e
in
co
r
r
ec
tly
id
en
tify
i
n
g
r
el
ev
an
t
in
s
tan
ce
s
.
I
n
ter
m
s
o
f
r
ec
all,
th
e
r
an
k
in
g
r
em
ain
s
co
n
s
is
ten
t,
with
th
e
RF
m
o
d
el
lead
in
g
th
e
lis
t
with
a
n
av
er
ag
e
o
f
9
3
%,
f
o
llo
wed
b
y
th
e
DT
m
o
d
el
an
d
SVM,
b
o
th
ac
h
iev
in
g
8
7
%.
T
h
is
in
d
icate
s
th
at
th
ese
m
o
d
els
h
av
e
a
s
tr
o
n
g
ab
ilit
y
to
p
r
ed
ict
b
o
th
tr
u
e
p
o
s
itiv
es
an
d
tr
u
e
n
e
g
ativ
es
ef
f
ec
tiv
ely
.
On
th
e
o
th
er
h
a
n
d
,
SGD,
LR
,
an
d
NB
ex
h
ib
it
m
ed
io
cr
e
s
c
o
r
es
in
r
ec
all,
with
v
alu
es
o
f
6
3
%,
5
0
%,
a
n
d
4
0
%,
r
esp
ec
tiv
ely
.
T
h
ese
lo
wer
s
co
r
es
s
u
g
g
est
a
r
elativ
ely
l
o
wer
ca
p
a
b
ilit
y
to
ac
cu
r
ately
p
r
e
d
ict
b
o
th
tr
u
e
p
o
s
itiv
e
an
d
f
alse
n
eg
ativ
e
in
s
tan
ce
s
f
o
r
th
ese
m
o
d
els.
I
n
te
r
m
s
o
f
F1
-
s
co
r
e,
t
h
e
o
r
d
er
o
f
p
e
r
f
o
r
m
a
n
ce
r
em
ai
n
s
co
n
s
is
ten
t,
with
th
e
RF
m
o
d
el
lead
in
g
at
9
3
%,
f
o
llo
wed
b
y
SVM
an
d
th
e
SGD
class
if
ier
,
b
o
th
ac
h
iev
in
g
8
7
%.
T
h
e
F1
-
s
co
r
e
is
a
h
ar
m
o
n
ic
m
ea
n
o
f
p
r
ec
is
io
n
an
d
r
ec
all,
r
ef
lectin
g
a
b
ala
n
ce
b
etwe
en
th
e
two
m
etr
ics.
T
h
e
r
est
o
f
th
e
m
o
d
els
ex
h
ib
it
F1
-
s
co
r
es
b
elo
w
6
4
%,
in
d
icati
n
g
a
lo
wer
b
alan
ce
b
etwe
en
p
r
ec
is
io
n
an
d
r
ec
all
f
o
r
th
ese
m
o
d
els
co
m
p
ar
ed
t
o
th
e
to
p
-
p
er
f
o
r
m
in
g
o
n
es.
As
ev
id
en
t
f
r
o
m
th
e
r
esu
lts
,
all
m
etr
ics
co
n
s
is
ten
tl
y
f
av
o
r
o
p
tim
is
tic
p
r
ed
ictio
n
s
f
o
r
RF
,
DT
,
an
d
SVM,
with
a
p
r
ef
er
en
ce
f
o
r
RF
.
T
h
is
s
u
g
g
ests
th
at
ML
ex
h
ib
its
th
e
ab
ilit
y
to
s
e
lect
m
ac
h
in
es
ev
en
with
lim
ited
d
ata.
Ho
w
ev
er
,
it'
s
im
p
o
r
tan
t
to
n
o
te
t
h
at
th
is
s
itu
atio
n
m
ay
in
tr
o
d
u
ce
b
ias.
T
h
is
will
b
e
f
u
r
th
er
i
n
v
esti
g
ated
in
t
h
is
p
ap
er
th
r
o
u
g
h
lear
n
in
g
cu
r
v
es
to
ass
ess
o
v
er
f
itti
n
g
in
m
o
d
els
with
g
o
o
d
p
r
ed
ictio
n
s
an
d
u
n
d
er
f
itti
n
g
in
m
o
d
els
with
lo
wer
s
co
r
es
[
2
9
]
.
4
.
2
.
Co
nfusi
o
n m
a
t
rix
T
h
e
co
n
f
u
s
io
n
m
at
r
ix
s
er
v
es
to
v
is
u
alize
b
o
th
tr
u
e
an
d
er
r
o
n
eo
u
s
p
r
ed
ictio
n
s
m
ad
e
b
y
ea
ch
m
ac
h
in
e
ac
r
o
s
s
d
if
f
er
en
t
class
es.
Fig
u
r
e
7
s
u
m
m
ar
izes
th
e
c
o
n
f
u
s
io
n
m
atr
ices
f
o
r
ea
c
h
m
o
d
el
tr
ain
ed
is
th
is
s
tu
d
y
,
p
r
o
v
id
i
n
g
a
co
m
p
r
eh
en
s
iv
e
o
v
er
v
iew
o
f
th
eir
p
er
f
o
r
m
a
n
ce
.
Ad
d
itio
n
ally
,
t
h
e
f
i
g
u
r
e
illu
s
tr
ates
th
e
d
is
tr
ib
u
tio
n
o
f
th
e
1
5
p
r
ed
icted
m
ac
h
in
es
ac
r
o
s
s
th
e
f
i
v
e
class
es,
o
f
f
er
in
g
a
v
is
u
al
r
ep
r
esen
tatio
n
o
f
h
o
w
well
ea
ch
m
o
d
el
class
if
ied
in
s
tan
ce
s
with
in
s
p
e
cif
ic
ca
teg
o
r
ies.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Ma
ch
in
e
lea
r
n
in
g
-
a
s
s
is
ted
d
ec
is
io
n
s
u
p
p
o
r
t in
in
d
u
s
tr
ia
l .
.
.
(
F
a
o
u
z
i Ta
ya
la
ti)
279
Fig
u
r
e
7
.
Ma
tr
ix
co
n
f
u
s
io
n
an
aly
s
is
f
o
r
th
e
s
ix
m
o
d
els
T
h
e
co
n
f
ir
m
atio
n
f
r
o
m
th
e
co
n
f
u
s
io
n
m
atr
ix
o
f
th
e
R
F
m
o
d
el
r
ein
f
o
r
ce
s
th
e
ac
cu
r
ac
y
o
f
p
r
ed
ictio
n
s
f
o
r
all
th
e
m
ac
h
in
es.
T
h
e
i
n
s
tan
ce
s
wer
e
ef
f
ec
tiv
ely
class
if
ied
an
d
d
is
tr
ib
u
ted
a
m
o
n
g
th
e
f
i
v
e
class
es,
in
d
icatin
g
th
e
m
o
d
el'
s
p
r
o
f
icie
n
cy
in
m
a
k
in
g
p
r
ec
is
e
p
r
e
d
ictio
n
s
f
o
r
e
ac
h
m
ac
h
i
n
e
ca
teg
o
r
y
.
Nev
er
t
h
eless
,
th
e
u
n
iq
u
e
in
s
tan
ce
o
f
in
ac
c
u
r
ate
p
r
ed
icti
o
n
o
cc
u
r
r
ed
with
in
class
4
(
m
ac
h
in
e
3
2
0
t)
,
wh
er
e
o
n
e
o
u
t
o
f
th
e
f
o
u
r
m
ac
h
in
es
was
er
r
o
n
eo
u
s
ly
ass
ig
n
ed
to
class
3
(
m
ac
h
in
e
2
5
0
t)
.
R
e
g
ar
d
in
g
SVM
an
d
DT
,
th
e
y
ex
h
ib
ited
id
en
tical
o
u
tco
m
es
b
y
co
r
r
ec
tly
p
r
ed
ict
in
g
class
es
1
,
2
,
an
d
5
.
Ho
we
v
er
,
b
o
th
m
o
d
els
m
ad
e
in
ac
cu
r
ate
p
r
e
d
ictio
n
s
f
o
r
o
n
e
m
ac
h
in
e
in
class
4
(
1
o
u
t
o
f
4
)
an
d
o
n
e
m
ac
h
in
e
in
clas
s
3
(
1
o
u
t
o
f
4
)
.
T
h
e
tr
ee
m
o
d
els
d
em
o
n
s
tr
ated
a
co
m
m
en
d
a
b
le
ab
ilit
y
to
d
is
tin
g
u
is
h
b
etwe
en
tr
u
e
p
o
s
itiv
es,
f
alse
n
eg
ativ
es,
an
d
f
alse
p
o
s
iti
v
es.
I
n
co
n
tr
ast,
th
e
NB
m
o
d
el
ex
h
ib
ited
in
co
r
r
ec
t p
r
ed
ictio
n
s
f
o
r
class
es
2
,
3
,
an
d
4
,
lea
d
in
g
t
o
f
ewe
r
tr
u
e
p
r
ed
ictio
n
s
(
n
eith
er
f
o
r
tr
u
e
p
o
s
itiv
es
n
o
r
f
alse
n
eg
ativ
es
)
.
C
o
n
s
eq
u
en
tly
,
th
is
im
p
ac
ted
th
e
r
ec
all,
wh
ich
s
tan
d
s
at
6
3
%.
T
h
e
NB
m
o
d
el
ap
p
ea
r
s
to
h
a
v
e
lim
ited
ca
p
a
b
ilit
y
in
d
is
tin
g
u
is
h
in
g
b
etwe
en
f
alse n
eg
ativ
es a
n
d
f
alse p
o
s
itiv
es.
4
.
3
.
L
ea
rning
curv
e
T
o
co
n
d
u
ct
a
co
m
p
r
eh
en
s
iv
e
ev
alu
atio
n
o
f
th
e
m
o
d
els
b
ased
o
n
th
eir
lear
n
in
g
ex
p
er
i
en
ce
s
,
we
em
p
lo
y
ed
lear
n
in
g
c
u
r
v
es
a
n
a
ly
s
is
[
3
0
]
as
s
h
o
wn
in
th
e
F
ig
u
r
e
8
.
T
h
is
a
p
p
r
o
ac
h
p
r
o
v
id
e
d
a
d
etailed
in
s
ig
h
t
in
to
h
o
w
ea
c
h
m
o
d
el
g
ain
e
d
k
n
o
wled
g
e
a
n
d
p
r
o
f
icien
c
y
o
v
e
r
tim
e.
L
ea
r
n
i
n
g
c
u
r
v
es ser
v
e
as a
wid
ely
ad
o
p
ted
d
iag
n
o
s
tic
to
o
l
in
th
e
f
ield
o
f
ML
,
p
ar
ticu
lar
ly
f
o
r
alg
o
r
ith
m
s
th
at
p
r
o
g
r
ess
iv
ely
ac
q
u
ir
e
k
n
o
wled
g
e
f
r
o
m
an
ex
p
an
d
i
n
g
t
r
ain
in
g
d
ataset.
I
n
th
is
p
a
p
er
,
th
e
ac
cu
r
ac
y
s
co
r
e
is
em
p
lo
y
e
d
as
m
et
r
ic
f
o
r
e
v
alu
atin
g
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
m
o
d
el
o
n
b
o
th
th
e
t
r
ain
in
g
s
et
an
d
th
e
v
alid
atio
n
s
et.
T
h
e
s
ix
alg
o
r
ith
m
s
ar
e
ca
teg
o
r
ized
b
ased
o
n
th
ei
r
tr
en
d
s
in
t
o
th
r
e
e
v
is
u
aliz
atio
n
s
f
o
r
an
al
y
s
is
.
R
eg
ar
d
in
g
th
e
Fig
u
r
e
8
(
a
)
,
it illu
s
tr
ates th
e
p
r
o
g
r
ess
io
n
o
f
NB
an
d
L
R
m
o
d
els.
B
o
th
m
o
d
els ex
h
ib
it a
s
im
ilar
ev
o
lu
tio
n
a
r
y
tr
en
d
,
s
h
o
wca
s
in
g
im
p
r
o
v
e
m
en
ts
in
p
er
f
o
r
m
an
ce
.
Ho
wev
er
,
th
ese
en
h
a
n
ce
m
en
ts
a
r
e
in
s
u
f
f
icien
t
to
m
atch
th
e
tr
ain
in
g
m
o
d
el,
lead
in
g
to
a
s
itu
atio
n
o
f
u
n
d
er
f
itti
n
g
.
T
h
e
m
o
d
e
l
s
tr
u
g
g
les
to
g
r
as
p
th
e
in
tr
icac
ies
o
f
th
e
tr
ain
in
g
d
ataset,
ev
i
d
en
t
in
th
e
tr
ain
i
n
g
c
u
r
v
e
m
ain
tain
in
g
a
n
ac
c
u
r
ac
y
ar
o
u
n
d
7
0
%.
E
x
am
in
in
g
th
e
lear
n
in
g
cu
r
v
e,
th
er
e
ex
is
ts
a
s
u
b
s
tan
tial
g
ap
b
etwe
en
th
e
tr
ain
in
g
an
d
v
alid
a
tio
n
s
ets
in
itial
ly
.
T
h
is
g
ap
g
r
ad
u
ally
d
im
in
is
h
e
s
with
th
e
ad
d
itio
n
o
f
tr
ain
in
g
ex
am
p
les,
s
tar
tin
g
f
r
o
m
a
d
ataset
s
ize
o
f
2
0
.
Nev
er
th
eless
,
th
e
ac
cu
r
ac
y
p
la
teau
s
at
6
0
%,
in
d
icatin
g
a
lim
itatio
n
in
th
e
m
o
d
el'
s
ab
ilit
y
to
g
en
er
alize
to
u
n
s
ee
n
d
ata.
E
v
en
wit
h
th
e
c
o
n
tin
u
o
u
s
ad
d
itio
n
o
f
tr
ain
in
g
ex
a
m
p
les,
th
er
e
is
n
o
d
is
ce
r
n
ib
le
im
p
r
o
v
e
m
en
t
i
n
p
er
f
o
r
m
an
ce
.
T
h
is
p
r
ed
icam
e
n
t
is
m
ir
r
o
r
ed
in
th
e
SGD
m
o
d
el
(
Fig
u
r
e
8
(
b
)
)
,
wh
er
e
th
e
tr
ain
in
g
m
o
d
el
f
ac
es
ch
allen
g
es in
lear
n
in
g
f
r
o
m
th
e
d
ataset,
r
esu
ltin
g
in
a
co
m
p
a
r
ab
le
in
ab
ilit
y
to
im
p
r
o
v
e
p
er
f
o
r
m
an
ce
.
Fig
u
r
e
8
(
c)
illu
s
tr
ates
th
e
p
r
o
g
r
ess
io
n
o
f
th
e
SVM,
R
F,
an
d
DT
m
o
d
els.
No
tab
ly
,
all
th
r
ee
ex
h
ib
it
a
s
im
ilar
tr
en
d
,
s
h
o
wca
s
in
g
co
n
s
is
ten
t
p
er
f
o
r
m
an
ce
im
p
r
o
v
e
m
en
t
o
v
e
r
tim
e.
T
h
is
im
p
lies
th
at
th
e
m
o
d
els
ar
e
u
n
d
er
g
o
in
g
ex
p
er
ien
tial
lear
n
i
n
g
,
e
v
id
en
t
with
in
th
e
in
ter
v
a
l
[
0
:2
0
]
.
T
h
e
ac
c
u
r
ac
y
is
at
it
s
m
in
im
u
m
f
o
r
th
e
m
o
d
els,
p
r
im
ar
ily
attr
ib
u
ted
to
th
e
lim
ited
tr
ain
in
g
ex
p
er
ie
n
ce
ac
ce
s
s
ib
le
d
u
r
in
g
th
is
p
h
a
s
e.
Ho
wev
er
,
with
in
th
e
r
an
g
e
[
2
0
:7
0
]
,
th
e
m
o
d
els
ex
h
ib
it
a
p
r
o
g
r
ess
iv
e
en
h
a
n
ce
m
en
t,
g
r
ad
u
ally
ap
p
r
o
ac
h
i
n
g
t
h
e
ac
cu
r
ac
y
o
f
th
e
tr
ain
in
g
s
et.
T
h
e
m
o
d
els
ca
n
g
en
er
ate
p
r
ed
ictio
n
s
f
o
r
n
o
v
el
d
ata
th
at
h
as
n
ev
er
b
ee
n
en
co
u
n
ter
ed
b
ef
o
r
e.
I
n
th
e
in
ter
v
al
[
7
0
:
1
0
0
]
,
b
o
t
h
th
e
R
F
an
d
SVM
ex
h
ib
it
a
co
m
b
in
at
io
n
o
f
v
ar
iatio
n
an
d
s
tab
ilit
y
.
T
h
is
s
u
g
g
ests
th
at
d
esp
ite
th
e
i
n
tr
o
d
u
ctio
n
o
f
n
ew
tr
ain
in
g
d
ata,
th
ese
two
m
o
d
els
ar
e
u
n
a
b
le
to
ac
h
ie
v
e
s
ig
n
if
ican
t
f
u
r
th
e
r
im
p
r
o
v
em
e
n
t.
Ho
we
v
er
,
i
n
th
e
ca
s
e
o
f
th
e
DT
m
o
d
el,
th
e
s
i
tu
atio
n
is
d
if
f
er
en
t.
I
t
co
n
tin
u
es
to
d
em
o
n
s
tr
ate
a
p
o
s
itiv
e
tr
en
d
o
f
im
p
r
o
v
em
e
n
t,
in
d
icatin
g
t
h
at
p
r
o
v
id
in
g
m
o
r
e
tr
ain
i
n
g
d
ata
c
o
u
ld
e
n
h
an
ce
its
lear
n
in
g
ex
p
er
ien
ce
a
n
d
p
er
f
o
r
m
an
ce
.
I
n
co
n
cl
u
s
io
n
,
co
n
s
id
er
in
g
th
e
t
r
ain
in
g
s
co
r
e'
s
h
ig
h
a
cc
u
r
ac
y
,
i
t
s
u
g
g
ests
lo
w
b
ias
Evaluation Warning : The document was created with Spire.PDF for Python.