I
nte
rna
t
io
na
l J
o
urna
l o
f
Adv
a
nces in Applie
d Science
s
(
I
J
AAS)
Vo
l.
14
,
No
.
2
,
J
u
n
e
2
0
2
5
,
p
p
.
384
~
3
9
3
I
SS
N:
2252
-
8
8
1
4
,
DOI
:
1
0
.
1
1
5
9
1
/ijaas
.
v
14
.
i
2
.
pp
384
-
3
9
3
384
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
a
s
.
ia
esco
r
e.
co
m
Co
mpa
ra
tive stu
dy
on fine
-
tuning
deep learning
mo
dels for
fruit
and v
eg
etabl
e clas
sifica
tion
Abd Ra
s
i
d M
a
m
a
t
1
,
M
o
ha
ma
d Af
endee
M
o
ha
m
ed
1
,
M
o
h
d
F
a
dzil Abd K
a
dir
1
,
No
rk
ha
ira
ni Abdu
l R
a
wi
2
,
A
zim
Z
a
lih
a
Abd Az
iz
2
,
Wa
n
Su
ry
a
ni Wa
n Aw
a
ng
1
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
F
a
c
u
l
t
y
o
f
I
n
f
o
r
ma
t
i
c
s
a
n
d
C
o
m
p
u
t
i
n
g
,
U
n
i
v
e
r
si
t
i
S
u
l
t
a
n
Za
i
n
a
l
A
b
i
d
i
n
,
T
e
r
e
n
g
g
a
n
u
,
M
a
l
a
y
s
i
a
2
D
e
p
a
r
t
me
n
t
o
f
M
u
l
t
i
me
d
i
a
S
t
u
d
i
e
s
,
F
a
c
u
l
t
y
o
f
I
n
f
o
r
ma
t
i
c
s
a
n
d
C
o
mp
u
t
i
n
g
,
U
n
i
v
e
r
si
t
i
S
u
l
t
a
n
Za
i
n
a
l
A
b
i
d
i
n
,
T
e
r
e
n
g
g
a
n
u
,
M
a
l
a
y
s
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l
21
,
2
0
2
4
R
ev
is
ed
Ap
r
1
,
2
0
2
5
Acc
ep
ted
Ap
r
23
,
2
0
2
5
F
ru
it
a
n
d
v
e
g
e
tab
le
re
c
o
g
n
it
i
o
n
a
n
d
c
las
sifica
ti
o
n
c
a
n
b
e
a
c
h
a
ll
e
n
g
in
g
tas
k
d
u
e
t
o
th
e
ir
d
i
v
e
rse
n
a
t
u
re
a
n
d
ha
ve
b
e
c
o
m
e
a
f
o
c
a
l
p
o
i
n
t
in
th
e
a
g
ricu
lt
u
ra
l
se
c
to
r.
In
a
d
d
it
i
o
n
t
o
th
a
t,
t
h
e
c
las
sifica
ti
o
n
o
f
fru
it
s
a
n
d
v
e
g
e
tab
les
in
c
re
a
se
s
th
e
c
o
st
o
f
lab
o
r
a
n
d
ti
m
e
.
In
re
c
e
n
t
y
e
a
rs,
d
e
e
p
lea
rn
in
g
a
p
p
li
c
a
ti
o
n
s
h
a
v
e
su
rg
e
d
t
o
th
e
fo
re
fr
o
n
t
,
o
ffe
ri
n
g
p
r
o
m
isin
g
so
l
u
ti
o
n
s.
P
a
rti
c
u
larly
,
t
h
e
c
las
sifica
ti
o
n
o
f
fr
u
it
s u
sin
g
ima
g
e
fe
a
tu
re
s
h
a
s g
a
rn
e
re
d
si
g
n
ifi
c
a
n
t
a
tt
e
n
ti
o
n
fro
m
re
se
a
rc
h
e
rs,
re
fl
e
c
ti
n
g
th
e
g
ro
win
g
imp
o
r
tan
c
e
o
f
th
is
a
r
e
a
in
th
e
a
g
ricu
lt
u
ra
l
d
o
m
a
in
.
In
t
h
is
wo
rk
,
t
h
e
fo
c
u
s
wa
s
o
n
fin
e
-
tu
n
in
g
h
y
p
e
r
p
a
ra
m
e
ters
a
n
d
th
e
e
v
a
lu
a
ti
o
n
o
f
a
sta
te
-
of
-
th
e
-
a
rt
d
e
e
p
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
two
r
k
(CNN
)
fo
r
t
h
e
c
las
sifica
ti
o
n
o
f
fr
u
it
s
a
n
d
v
e
g
e
tab
les
.
Am
o
n
g
th
e
h
y
p
e
rp
a
ra
m
e
ters
stu
d
ie
d
a
re
t
h
e
n
u
m
b
e
r
o
f
b
a
tch
siz
e
,
n
u
m
b
e
r
o
f
e
p
o
c
h
s,
ty
p
e
o
f
o
p
t
imiz
e
r,
re
c
ti
fied
u
n
i
t,
a
n
d
d
ro
p
o
u
t.
Th
e
d
a
tas
e
t
u
s
e
d
is
t
h
e
f
ru
it
_
v
e
g
e
tab
le
d
a
tas
e
t
wh
ich
c
o
n
sists
o
f
3
6
c
las
se
s
a
n
d
e
a
c
h
c
las
s
c
o
n
tai
n
s
1
,
0
0
0
ima
g
e
s.
Th
e
re
su
lt
s
sh
o
w
th
a
t
th
e
p
ro
p
o
se
d
m
o
d
e
l
b
a
se
d
o
n
th
e
b
a
tc
h
siz
e
=
6
4
a
n
d
th
e
n
u
m
b
e
r
o
f
e
p
o
c
h
s=
2
5
,
p
r
o
d
u
c
e
s
th
e
m
o
st
o
p
t
i
m
a
l
m
o
d
e
l
with
a
n
a
c
c
u
ra
c
y
v
a
l
u
e
(trai
n
in
g
)
o
f
9
9
.
0
2
%
,
wh
il
e
t
h
e
v
a
li
d
a
ti
o
n
is
9
5
.
7
3
%
a
n
d
t
h
e
lo
ss
is
6
.
0
6
%
(m
in
imu
m
).
K
ey
w
o
r
d
s
:
Acc
u
r
ac
y
C
las
s
if
icatio
n
C
o
n
v
o
lu
tio
n
n
eu
r
al
n
etwo
r
k
Hy
p
er
p
ar
a
m
eter
L
o
s
s
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
:
Ab
d
R
asid
Ma
m
at
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
Facu
lty
o
f
I
n
f
o
r
m
atics a
n
d
C
o
m
p
u
tin
g
,
Un
iv
er
s
iti
Su
ltan
Z
ain
al
Ab
id
in
B
esu
t Cam
p
u
s
,
2
2
2
0
0
B
esu
t
,
T
er
en
g
g
a
n
u
Dar
u
l I
m
an
,
Ma
lay
s
ia
E
m
ail: a
r
m
@
u
n
is
za
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
I
m
a
g
e
c
l
a
s
s
i
f
i
c
a
ti
o
n
h
a
s
it
s
r
o
o
t
s
i
n
a
wi
d
e
r
a
n
g
e
o
f
t
ec
h
n
o
l
o
g
i
e
s
d
e
v
e
l
o
p
e
d
o
v
e
r
m
a
n
y
y
e
a
r
s
[
1
]
–
[
3
]
.
T
h
e
s
e
t
e
c
h
n
o
l
o
g
i
es
i
n
c
l
u
d
e
a
d
v
a
n
c
e
m
e
n
ts
i
n
c
o
m
p
u
t
e
r
v
i
s
i
o
n
,
m
a
c
h
i
n
e
le
a
r
n
i
n
g
,
a
n
d
p
a
t
t
e
r
n
r
e
c
o
g
n
i
t
i
o
n
.
R
es
e
a
r
c
h
e
r
s
a
n
d
en
g
i
n
e
e
r
s
h
a
v
e
w
o
r
k
e
d
e
x
te
n
s
i
v
e
l
y
t
o
i
m
p
r
o
v
e
t
h
e
a
b
i
l
it
y
o
f
m
a
c
h
i
n
e
s
t
o
r
e
c
o
g
n
i
z
e
a
n
d
c
a
t
e
g
o
r
i
z
e
i
m
a
g
es
a
c
c
u
r
at
e
l
y
.
A
s
a
r
es
u
l
t
o
f
t
h
es
e
e
f
f
o
r
t
s
,
i
m
a
g
e
c
l
ass
i
f
i
c
at
i
o
n
s
y
s
te
m
s
h
a
v
e
b
e
c
o
m
e
i
n
c
r
e
d
i
b
l
y
s
o
p
h
i
s
t
i
c
at
e
d
a
n
d
p
o
w
e
r
f
u
l
.
T
h
e
a
p
p
l
i
c
a
ti
o
n
s
o
f
th
e
s
e
c
l
ass
i
f
i
e
r
s
a
r
e
b
o
t
h
v
a
r
i
ed
a
n
d
w
i
d
e
l
y
u
s
e
d
a
c
r
o
s
s
m
a
n
y
d
i
f
f
e
r
e
n
t
f
i
el
d
s
.
So
m
e
a
p
p
l
i
c
a
ti
o
n
s
f
o
c
u
s
o
n
b
r
o
a
d
t
a
s
k
s
,
s
u
c
h
as
i
n
t
e
r
p
r
e
t
i
n
g
a
n
d
u
n
d
e
r
s
t
a
n
d
i
n
g
t
h
e
o
v
e
r
a
l
l
c
o
n
t
e
x
t o
f
a
n
i
m
a
g
e
.
O
th
e
r
s
a
r
e
d
e
s
i
g
n
e
d
f
o
r
h
i
g
h
l
y
s
p
e
c
i
f
i
c
a
n
d
c
r
it
i
c
a
l p
u
r
p
o
s
es
,
li
k
e
d
e
t
e
ct
i
n
g
s
i
g
n
s
o
f
c
a
n
c
e
r
i
n
p
a
t
i
e
n
t
t
e
s
t
r
e
s
u
lt
s
[
2
]
,
[
4
]
,
[
5
]
.
O
v
e
r
a
l
l
,
t
h
e
i
m
p
a
c
t
o
f
i
m
a
g
e
c
l
a
s
s
i
f
ic
a
t
i
o
n
t
ec
h
n
o
l
o
g
i
e
s
c
o
n
t
i
n
u
es
t
o
g
r
o
w
,
i
n
f
l
u
e
n
c
i
n
g
i
n
d
u
s
t
r
i
es
f
r
o
m
a
g
r
i
c
u
l
t
u
r
e
,
t
r
a
n
s
p
o
r
t
at
i
o
n
,
a
n
d
m
i
l
i
t
a
r
y
t
o
n
a
m
e
a
f
e
w
.
R
esear
ch
in
f
r
u
it
p
r
o
ce
s
s
in
g
h
o
ld
s
s
u
b
s
tan
tial
s
ig
n
if
ican
ce
wh
en
co
n
n
ec
ted
with
o
th
er
ec
o
n
o
m
ic
s
ec
to
r
s
lik
e
ag
r
icu
ltu
r
e,
e
n
co
m
p
ass
in
g
b
o
th
wh
o
lesale
an
d
r
etail
m
ar
k
ets,
alo
n
g
s
id
e
th
e
p
r
o
ce
s
s
in
g
in
d
u
s
tr
y
.
T
h
ese
f
ac
to
r
s
h
av
e
p
u
s
h
e
d
r
es
ea
r
ch
er
s
wh
o
d
ev
elo
p
ed
v
ar
io
u
s
m
eth
o
d
s
to
au
to
m
atica
lly
p
r
o
ce
s
s
f
r
u
its
,
eith
er
to
id
en
tify
o
r
esti
m
ate
th
e
ir
q
u
ality
ef
f
icien
tly
,
class
if
icatio
n
,
an
d
p
ac
k
a
g
in
g
to
war
d
s
m
o
r
e
f
o
cu
s
ed
r
esear
ch
.
T
h
is
is
b
ec
au
s
e
th
e
p
r
o
ce
s
s
in
g
,
class
if
icatio
n
,
an
d
s
ep
ar
atio
n
o
f
p
lan
ts
r
e
q
u
ir
e
in
te
n
s
iv
e
en
er
g
y
an
d
tim
e
[
6
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
C
o
mp
a
r
a
tive
s
tu
d
y
o
n
fin
e
-
t
u
n
in
g
d
ee
p
lea
r
n
in
g
mo
d
els fo
r
fr
u
it a
n
d
ve
g
eta
b
le
…
(
A
b
d
R
a
s
id
Ma
ma
t)
385
T
h
e
m
o
s
t
ef
f
icien
t
class
if
ier
s
th
at
p
e
r
f
o
r
m
well
i
n
class
if
y
in
g
im
ag
es
s
u
ch
as
f
r
u
its
ar
e
th
o
s
e
b
y
u
s
in
g
d
ee
p
lear
n
i
n
g
alg
o
r
ith
m
s
[
7
]
–
[
9
]
.
Dee
p
lear
n
in
g
,
alth
o
u
g
h
a
th
e
o
r
etica
l
co
n
ce
p
t,
is
n
o
th
in
g
n
ew.
I
t
h
as
en
jo
y
ed
a
s
u
r
g
e
o
f
in
ter
est
o
v
er
th
e
p
ast
d
ec
ad
e,
b
ec
o
m
in
g
t
h
e
h
o
ttes
t
tr
en
d
in
m
ac
h
in
e
le
ar
n
in
g
d
u
e
to
m
an
y
f
ac
to
r
s
.
Dee
p
lear
n
in
g
ap
p
r
o
a
ch
es
h
av
e
s
ig
n
if
ican
tly
o
u
tp
e
r
f
o
r
m
ed
s
tate
-
of
-
th
e
-
ar
t
ap
p
r
o
a
ch
es
in
m
an
y
task
s
ac
r
o
s
s
d
if
f
er
en
t
f
ield
s
s
u
ch
as
d
ata
in
tr
u
s
io
n
,
im
a
g
e
class
if
icatio
n
,
co
m
p
u
ter
v
is
io
n
,
s
p
ee
ch
p
r
o
ce
s
s
in
g
,
a
n
d
n
atu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
(
NL
P)
[
1
0
]
–
[
1
2
]
.
Ho
wev
e
r
,
th
er
e
r
em
ai
n
s
an
o
p
p
o
r
t
u
n
ity
to
ex
p
lo
r
e
h
y
p
er
p
ar
am
eter
s
i
n
d
ee
p
lear
n
in
g
t
o
ac
h
iev
e
th
e
b
est
class
if
icatio
n
r
esu
lts
th
r
o
u
g
h
f
in
e
-
t
u
n
in
g
.
T
h
e
f
o
cu
s
is
o
n
th
e
b
atch
s
ize,
th
e
n
u
m
b
e
r
o
f
e
p
o
ch
s
,
an
d
th
eir
r
elatio
n
s
h
ip
,
wh
ile
th
e
o
th
er
h
y
p
er
p
a
r
am
eter
s
ar
e
s
et
as
o
u
tlin
ed
in
s
ec
tio
n
2
.
2
.
Sev
er
al
s
tu
d
ies
r
elate
d
to
au
to
m
atic
f
r
u
it
class
if
icatio
n
h
av
e
b
ee
n
d
o
n
e
s
o
f
ar
b
y
m
an
y
s
cien
tis
ts
an
d
r
esear
ch
er
s
.
I
t
in
v
o
lv
es
f
r
u
it
class
if
icatio
n
,
d
eter
m
in
in
g
r
ip
en
ess
s
tag
e,
d
is
ea
s
e
d
etec
ti
o
n
,
ass
ess
in
g
citr
u
s
lev
el
,
an
d
class
if
y
in
g
f
r
esh
f
r
u
it
p
alm
o
il
in
r
ip
e
b
u
n
c
h
es
[
1
3
]
–
[
1
5
]
.
Fo
r
ex
am
p
le,
f
o
r
th
e
p
r
o
ce
s
s
o
f
h
ar
v
esti
n
g
d
ates
b
ased
o
n
5
d
if
f
er
en
t
class
if
icatio
n
s
o
f
d
ates
,
a
r
o
b
o
tic
h
ar
v
esti
n
g
m
o
d
e
l
was
p
r
o
p
o
s
ed
b
y
Altah
er
i
et
a
l.
[
1
6
]
.
T
h
e
m
o
d
e
l
u
s
es
an
in
ter
n
al
d
ataset
co
n
tain
in
g
8
,
0
0
0
im
ag
es
f
o
r
tr
ain
i
n
g
an
d
test
in
g
an
d
ac
h
iev
es 9
9
% a
cc
u
r
ac
y
.
Oth
er
r
esear
ch
er
s
in
[
1
7
]
,
d
e
v
elo
p
ed
a
f
r
u
i
t
class
if
icatio
n
m
o
d
el
f
o
r
in
d
u
s
tr
ial
ap
p
licati
o
n
s
.
I
n
th
is
m
o
d
el,
th
e
a
u
th
o
r
s
u
s
ed
p
u
b
lic
d
atasets
an
d
o
n
e
o
f
th
e
d
atasets
co
n
tain
ed
im
ag
es
o
f
f
r
u
its
wh
ich
ar
e
co
m
p
le
x
to
id
en
tify
.
T
h
e
ac
cu
r
ac
y
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el
is
8
5
%.
Nex
t,
th
e
au
th
o
r
s
in
th
eir
r
ese
ar
ch
em
p
lo
y
e
d
co
n
v
o
l
u
tio
n
al
n
e
u
r
al
n
etwo
r
k
s
(
C
NNs)
to
class
if
y
f
r
u
its
an
d
v
e
g
etab
les
b
ased
o
n
R
GB
im
ag
e
d
ata
[
1
8
]
.
Desp
ite
th
e
ch
allen
g
e
o
f
s
o
m
e
item
s
s
h
ar
in
g
s
im
ilar
co
l
o
r
s
an
d
s
h
ap
es,
th
ei
r
ap
p
r
o
ac
h
ac
h
iev
ed
im
p
r
o
v
ed
class
if
icatio
n
ac
cu
r
ac
y
co
m
p
a
r
ed
to
o
t
h
er
m
eth
o
d
s
.
Dee
p
lear
n
in
g
m
eth
o
d
s
f
o
r
f
r
u
it
class
if
icatio
n
ar
e
ex
ten
s
iv
ely
ap
p
lied
in
t
h
e
p
o
s
t
-
h
ar
v
es
t
s
tag
e
an
d
th
e
f
r
u
it
in
d
u
s
tr
y
.
I
n
a
p
a
r
ticu
lar
s
tu
d
y
,
a
C
NN
-
b
ased
m
o
d
el
was
in
tr
o
d
u
ce
d
to
ca
teg
o
r
ize
ap
p
les
in
to
n
o
r
m
al
an
d
d
ef
ec
tiv
e
class
es.
T
h
is
m
o
d
el
was
in
te
g
r
ated
in
to
a
f
r
u
it
s
o
r
tin
g
s
y
s
tem
,
d
em
o
n
s
tr
atin
g
a
r
em
ar
k
ab
le
ac
cu
r
ac
y
o
f
9
2
%
wh
ile
p
r
o
c
ess
in
g
ea
ch
ap
p
le
in
less
th
an
7
2
m
illi
s
ec
o
n
d
s
[
1
9
]
.
Desp
ite
th
e
in
cr
ea
s
in
g
in
ter
est
in
ar
tific
ial
in
tellig
e
n
ce
(
AI
)
f
o
r
r
ed
u
cin
g
f
o
o
d
waste,
th
er
e
r
em
ain
s
a
n
o
ta
b
le
g
ap
in
r
esear
ch
co
n
ce
r
n
in
g
th
e
a
p
p
licatio
n
o
f
AI
f
o
r
class
if
y
in
g
an
d
d
etec
tin
g
f
r
u
its
an
d
v
eg
eta
b
les.
T
h
i
s
s
tu
d
y
aim
s
to
f
ill
th
is
g
ap
b
y
d
esig
n
in
g
a
n
d
im
p
lem
en
tin
g
an
in
tellig
en
t
s
y
s
tem
ca
p
ab
le
o
f
ac
cu
r
ately
id
e
n
tify
in
g
3
6
d
i
f
f
er
en
t
class
es o
f
f
r
u
its
an
d
v
eg
etab
le
s
th
r
o
u
g
h
th
e
tu
n
i
n
g
o
f
h
y
p
er
p
ar
am
eter
s
.
2.
RE
S
E
ARCH
M
E
T
HOD
Dee
p
lear
n
in
g
is
a
h
ig
h
l
y
ac
ti
v
e
r
esear
ch
ar
ea
in
c
o
m
p
u
te
r
v
is
io
n
an
d
im
ag
e
class
if
icatio
n
.
A
ty
p
ical
ar
ch
itectu
r
e
o
f
d
ee
p
C
NN
co
m
p
r
is
es
an
in
p
u
t
lay
e
r
,
an
o
u
tp
u
t
o
r
class
if
icatio
n
lay
er
,
a
n
d
m
u
ltip
le
h
id
d
en
lay
er
s
(
f
ea
tu
r
e
ex
tr
ac
tio
n
is
d
o
n
e
in
th
e
h
i
d
d
en
lay
e
r
s
)
.
T
h
ese
h
id
d
en
lay
er
s
o
f
te
n
in
clu
d
e
co
n
v
o
lu
tio
n
al,
p
o
o
lin
g
,
an
d
f
u
lly
co
n
n
ec
ted
l
ay
er
s
,
alo
n
g
with
th
e
p
o
s
s
ib
ilit
y
o
f
a
S
o
f
t
M
ax
la
y
er
[
2
0
]
–
[
2
2
]
.
I
n
g
e
n
er
al,
th
e
h
y
p
er
p
ar
am
eter
s
ar
e
f
ilter
s
ize
(
k
er
n
el
s
ize)
,
n
u
m
b
er
o
f
f
il
ter
s
,
p
o
o
lin
g
,
ac
tiv
atio
n
f
u
n
ct
io
n
,
lear
n
i
n
g
r
ate,
b
atch
s
ize,
n
u
m
b
er
o
f
ep
o
c
h
s
,
an
d
d
r
o
p
o
u
t
[
2
2
]
,
[
2
3
]
.
Su
b
s
eq
u
en
tly
,
th
e
tu
n
in
g
o
r
ad
ju
s
tm
en
t
o
f
th
e
h
y
p
er
p
ar
a
m
eter
s
is
b
atch
s
ize
an
d
th
e
n
u
m
b
er
o
f
ep
o
ch
s
,
wh
ile
o
t
h
er
p
a
r
am
ete
r
s
,
n
am
ely
th
e
ac
tiv
atio
n
f
u
n
c
tio
n
,
d
r
o
p
o
u
t
v
alu
e,
lear
n
in
g
r
ate,
an
d
o
p
tim
izer
ty
p
e,
ar
e
k
e
p
t
f
ix
e
d
.
Sp
ec
if
ica
lly
,
t
h
e
r
ec
tifie
d
l
in
ea
r
u
n
it
(
R
eL
U)
is
u
tili
ze
d
as
th
e
ac
tiv
ati
o
n
f
u
n
ctio
n
,
Ad
am
s
er
v
es
as
th
e
o
p
tim
izer
,
a
n
d
le
ar
n
in
g
r
ate
is
an
au
to
m
atic
m
eth
o
d
a
n
d
a
d
r
o
p
o
u
t
r
ate
o
f
0
.
2
(
2
0
%)
is
ap
p
lied
.
On
th
e
o
th
er
h
a
n
d
,
th
e
v
ar
iab
l
es
b
atch
s
ize
an
d
n
u
m
b
e
r
o
f
e
p
o
ch
s
ar
e
v
ar
ia
b
le
p
ar
am
eter
s
th
at
will
f
in
e
-
tu
n
e
th
e
m
o
d
el
tr
ain
i
n
g
p
r
o
ce
s
s
f
o
r
f
in
d
in
g
a
n
d
o
p
tim
izin
g
th
e
ac
c
u
r
ac
y
p
e
r
f
o
r
m
an
ce
f
o
r
th
e
d
at
a
s
et.
2
.
1
.
Da
t
a
s
et
Sam
p
le
im
ag
es
co
n
s
is
tin
g
o
f
r
ea
l
-
wo
r
l
d
in
f
o
r
m
atio
n
ar
e
r
ef
er
r
ed
to
as
d
ataset
s
an
d
d
ig
ital
im
ag
e
co
llectio
n
is
k
n
o
wn
as
d
ata
ac
q
u
is
itio
n
[
2
4
]
.
T
h
e
d
ataset
u
s
ed
to
w
o
r
k
is
th
e
f
r
u
it_
v
e
g
etab
le
d
ataset
wh
ich
is
p
u
b
licly
av
aila
b
le
o
n
Kag
g
le
wh
ich
is
p
u
b
licly
av
ailab
le.
I
t
is
f
r
ee
a
n
d
d
o
wn
lo
a
de
d
f
r
o
m
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
krit
ikseth
/f
r
u
it
-
and
-
ve
g
eta
b
le
-
ima
g
e
-
r
ec
o
g
n
itio
n
.
Ma
n
y
r
esear
ch
er
s
u
s
ed
th
is
d
ataset
f
o
r
th
eir
r
esear
ch
[
2
5
]
,
[
2
6
]
.
Fig
u
r
e
1
s
h
o
ws th
e
e
x
am
p
le
o
f
im
ag
es in
th
e
d
ataset.
T
h
is
d
ataset
co
n
tain
s
3
6
class
es
an
d
1
,
0
0
0
im
ag
es
f
o
r
ea
ch
class
.
Fig
u
r
e
1
,
s
h
o
ws
an
ex
am
p
le
o
f
th
e
d
if
f
er
en
t
class
es
o
f
f
r
u
its
we
u
s
ed
f
o
r
an
al
y
s
is
.
T
h
e
n
th
e
d
ataset
is
s
p
lit
in
to
3
ca
teg
o
r
ies
n
am
ely
tr
ain
in
g
d
ataset
(
8
0
%),
test
in
g
d
ataset
(
1
0
%)
,
an
d
f
in
ally
v
alid
atio
n
d
ataset
(
1
0
%)
[
8
]
,
[
2
7
]
.
2
.
1
.
1
.
T
ra
ini
ng
da
t
a
s
et
E
ac
h
class
in
th
e
tr
ain
in
g
d
ataset
co
n
tain
s
1
0
0
im
ag
es,
p
r
o
v
id
in
g
a
b
alan
ce
d
r
ep
r
esen
tatio
n
f
o
r
ea
ch
ca
teg
o
r
y
.
Alto
g
eth
e
r
,
th
e
d
ata
s
et
is
m
ad
e
u
p
o
f
a
to
tal
o
f
3
,
6
0
0
im
ag
es,
en
s
u
r
in
g
a
s
u
b
s
tan
tial
am
o
u
n
t
o
f
d
ata
f
o
r
th
e
lear
n
in
g
p
r
o
ce
s
s
.
T
h
is
tr
ain
in
g
d
ata
s
et
p
lay
s
a
cr
u
cial
r
o
le
in
d
ev
elo
p
in
g
th
e
n
etwo
r
k
m
o
d
el
b
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
14
,
No
.
2
,
J
u
n
e
2
0
2
5
:
3
8
4
-
393
386
p
r
o
v
id
i
n
g
th
e
e
x
am
p
les
n
ee
d
ed
f
o
r
lea
r
n
in
g
.
I
t
is
s
p
ec
if
ic
ally
u
s
ed
to
tr
ai
n
th
e
p
a
r
am
e
ter
s
o
f
th
e
m
o
d
el,
in
clu
d
in
g
a
d
ju
s
tin
g
th
e
weig
h
ts
an
d
b
iases
to
o
p
tim
ize
p
e
r
f
o
r
m
an
ce
.
2
.
1
.
2
.
Va
lid
a
t
io
n da
t
a
s
et
T
h
e
v
alid
atio
n
d
ataset
is
u
s
e
d
to
ev
alu
ate
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
tr
ain
e
d
m
o
d
el
an
d
en
s
u
r
e
it
g
en
er
alize
s
well
to
n
ew
d
ata.
Fo
r
th
is
p
ar
ticu
lar
task
,
ea
ch
c
lass
in
th
e
d
ataset
co
n
tain
s
1
0
im
ag
es,
p
r
o
v
i
d
in
g
a
s
m
aller
b
u
t
b
alan
c
ed
s
et
f
o
r
ass
ess
m
en
t.
I
n
to
tal,
th
e
v
a
lid
atio
n
d
ataset
co
n
s
is
ts
o
f
3
6
0
im
a
g
es,
wh
ich
r
ep
r
esen
ts
1
0
% o
f
th
e
en
tire
d
ataset.
T
h
is
v
alid
atio
n
s
tep
is
ess
en
tial
f
o
r
m
o
n
it
o
r
in
g
th
e
m
o
d
el’
s
ac
cu
r
ac
y
an
d
p
r
ev
en
tin
g
is
s
u
es su
ch
as o
v
er
f
itti
n
g
.
2
.
1
.
3
.
T
esting
da
t
a
s
et
T
h
e
aim
o
f
u
s
in
g
a
test
in
g
d
ataset
in
d
ee
p
lear
n
in
g
is
to
p
r
o
v
i
d
e
an
in
d
e
p
en
d
e
n
t
an
d
u
n
b
iased
ev
alu
atio
n
o
f
a
tr
ain
e
d
m
o
d
el’
s
p
er
f
o
r
m
a
n
ce
.
T
h
is
d
ataset
is
cr
u
cial
b
ec
au
s
e
it
h
el
p
s
d
eter
m
in
e
h
o
w
well
th
e
m
o
d
el
ca
n
m
ak
e
p
r
ed
ictio
n
s
o
n
p
r
ev
io
u
s
ly
u
n
s
ee
n
d
ata.
F
o
r
th
is
ev
alu
atio
n
p
r
o
ce
s
s
,
a
t
o
tal
o
f
3
6
0
i
m
ag
es
wer
e
u
s
ed
,
e
n
s
u
r
in
g
a
f
air
a
n
d
r
ep
r
esen
tativ
e
test
.
T
h
ese
im
a
g
es
wer
e
ev
e
n
ly
ta
k
en
f
r
o
m
te
n
d
if
f
er
en
t
class
es,
m
ain
tain
in
g
b
ala
n
ce
ac
r
o
s
s
th
e
ca
teg
o
r
ies.
Fig
u
r
e
1
.
An
ex
am
p
le
o
f
im
a
g
es in
a
d
ataset
2
.
2
.
T
un
ing
hy
per
pa
ra
m
et
e
rs o
f
t
he
pro
po
s
ed
m
o
dels
I
n
th
is
wo
r
k
,
we
h
a
v
e
in
c
o
r
p
o
r
ated
m
o
d
el
-
tu
n
i
n
g
tec
h
n
iq
u
e
s
to
p
r
e
v
en
t
t
h
e
m
o
d
el
f
r
o
m
o
v
er
f
itti
n
g
.
T
h
e
h
y
p
er
p
ar
am
eter
s
tu
n
ed
ar
e
th
e
b
atch
s
ize
an
d
n
u
m
b
er
o
f
e
p
o
ch
s
,
with
a
o
n
th
eir
r
elatio
n
s
h
ip
an
d
s
u
m
m
ar
ized
as f
o
llo
ws
:
i)
B
atch
-
s
ize
,
t
h
e
b
atch
s
ize
r
ef
e
r
s
to
th
e
n
u
m
b
er
o
f
tr
ai
n
in
g
e
x
am
p
les
u
tili
ze
d
in
o
n
e
iter
ati
o
n
in
tr
ain
in
g
n
eu
r
al
n
etwo
r
k
s
.
Nu
m
b
e
r
b
atc
h
s
ize
is
u
s
ed
1
6
,
3
2
,
an
d
6
4
.
ii)
Nu
m
b
er
o
f
e
p
o
c
h
s
,
t
h
e
n
u
m
b
er
o
f
ep
o
ch
s
b
ased
o
n
th
e
n
u
m
b
er
o
f
tim
es
th
e
en
tire
d
ata
s
et
is
p
ass
ed
f
o
r
war
d
a
n
d
b
ac
k
war
d
th
r
o
u
g
h
th
e
n
eu
r
al
n
etwo
r
k
d
u
r
in
g
th
e
tr
ain
in
g
p
r
o
ce
s
s
.
T
h
e
p
er
f
o
r
m
an
ce
is
co
m
p
ar
ed
o
n
t
h
e
ep
o
c
h
s
n
u
m
b
er
1
0
,
15
,
a
n
d
2
5
.
iii)
Op
tim
izer
,
Ad
am
is
ad
ap
tiv
e
alg
o
r
ith
m
an
d
it
is
u
s
ed
to
o
p
tim
ize
all
th
e
ex
p
er
im
en
ts
.
T
y
p
ically
,
wh
en
u
s
in
g
Ad
am
,
th
e
lear
n
in
g
r
ate
r
an
g
es f
r
o
m
0
.
0
0
0
1
to
0
.
0
0
1
,
with
a
s
tar
tin
g
p
o
in
t
o
f
0
.
0
0
1
[
2
7
]
,
[
2
8
]
.
iv
)
R
eL
U
is
a
p
o
p
u
lar
an
d
c
o
m
m
o
n
ac
tiv
atio
n
f
u
n
ctio
n
u
s
ed
in
d
ee
p
lear
n
in
g
[
2
9
]
–
[
3
1
]
.
T
h
e
m
ath
em
atica
l
ex
p
r
ess
io
n
f
o
r
R
eL
U
is
s
h
o
wn
in
(
1
)
.
(
)
=
(
0
,
)
(
1
)
W
h
er
e
x
is
th
e
in
p
u
t t
o
th
e
ac
t
iv
atio
n
f
u
n
ctio
n
an
d
f
(
x
)
is
th
e
o
u
tp
u
t
a
f
ter
ap
p
ly
in
g
R
eL
U
a
ctiv
atio
n
.
v)
T
h
e
d
r
o
p
o
u
t te
ch
n
iq
u
e
h
el
p
s
av
o
id
th
e
is
s
u
e
o
f
o
v
e
r
f
itti
n
g
an
d
d
u
r
in
g
th
e
tr
ain
in
g
,
n
eu
r
o
n
s
ar
e
r
an
d
o
m
ly
ch
o
s
en
an
d
d
is
ca
r
d
ed
.
I
n
t
h
is
m
o
d
el
,
th
e
d
r
o
p
o
u
t
v
alu
e
is
0
.
2
[
3
2
]
,
[
3
3
]
.
2
.
3
.
P
r
o
po
s
ed
m
o
del
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
is
co
n
s
id
er
ed
o
n
e
o
f
th
e
m
o
s
t
co
m
m
o
n
d
ee
p
lear
n
in
g
ar
ch
itectu
r
es
[
2
1
]
,
[
3
4
]
,
as
s
h
o
wn
in
Fig
u
r
e
2
.
T
h
e
n
e
two
r
k
ar
c
h
itectu
r
e
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el
is
d
esig
n
ed
as
a
s
eq
u
en
ce
m
o
d
el
to
h
an
d
le
s
eq
u
e
n
tial d
ata.
A
s
u
m
m
ar
y
o
f
th
e
m
o
d
el
is
as f
o
llo
w
s
:
i)
Firstl
y
,
th
e
in
p
u
t
lay
er
.
T
h
is
l
ay
er
r
e
p
r
esen
ts
th
e
r
aw
in
p
u
t
d
ata,
s
u
ch
as
an
im
ag
e.
E
ac
h
p
ix
el
in
th
e
im
ag
e
m
ay
b
e
r
ep
r
esen
ted
as
a
s
ep
ar
ate
in
p
u
t
n
o
d
e.
I
t
is
f
o
ll
o
wed
b
y
a
p
r
e
p
r
o
ce
s
s
in
g
la
y
er
as
a
f
u
n
ctio
n
to
n
o
r
m
alize
th
e
p
ix
el
o
f
im
ag
es
in
to
r
an
g
e
0
-
1
[
3
5
]
.
I
t
is
a
g
o
o
d
p
r
ac
tice
to
n
o
r
m
alize
th
e
d
ata
to
av
o
id
d
if
f
er
en
t scale
s
o
f
th
e
f
ea
tu
r
e
v
ec
to
r
s
an
d
t
h
u
s
im
p
r
o
v
e
d
ata
in
teg
r
ity
[
3
6
]
,
[
3
7
]
.
ii)
T
h
e
n
ex
t la
y
er
is
th
e
C
o
n
v
2
D
(
1
6
,
3
,
R
eL
U)
lay
er
,
wh
er
e
th
is
lay
er
co
n
tain
s
1
6
f
ilter
s
(
o
r
k
er
n
els)
o
f
s
ize
3
×
3
an
d
th
e
ac
tiv
atio
n
ar
e
R
eL
U.
E
ac
h
f
ilter
lear
n
s
d
if
f
er
en
t f
ea
tu
r
es f
r
o
m
th
e
in
p
u
t d
ata.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
C
o
mp
a
r
a
tive
s
tu
d
y
o
n
fin
e
-
t
u
n
in
g
d
ee
p
lea
r
n
in
g
mo
d
els fo
r
fr
u
it a
n
d
ve
g
eta
b
le
…
(
A
b
d
R
a
s
id
Ma
ma
t)
387
iii)
T
h
e
Ma
x
Po
o
lin
g
2
D
lay
er
is
a
p
o
o
lin
g
o
p
er
atio
n
ty
p
ically
u
s
ed
i
n
C
NNs
f
o
r
d
o
wn
-
s
am
p
lin
g
f
ea
tu
r
e
m
ap
s
.
I
t
h
elp
s
r
e
d
u
ce
t
h
e
s
p
a
tial
d
im
en
s
io
n
s
o
f
th
e
in
p
u
t,
t
h
er
eb
y
r
ed
u
cin
g
co
m
p
u
tatio
n
al
co
m
p
lex
ity
an
d
co
n
tr
o
llin
g
o
v
e
r
f
itti
n
g
.
T
h
e
n
ex
t
lay
e
r
is
C
o
n
v
2
D
(
3
2
,
3
R
eL
U)
.
I
n
th
is
ca
s
e,
th
e
r
e
ar
e
3
2
f
ilter
s
an
d
th
e
s
ize
is
3
×
3
R
eL
U
is
u
s
ed
f
o
r
th
e
ac
tiv
atio
n
f
u
n
ctio
n
.
iv
)
T
h
e
f
o
llo
win
g
lay
er
is
C
o
n
v
2
D
(
6
4
,
3
R
eL
U)
.
T
h
is
lay
e
r
s
h
o
ws
6
4
f
ilt
er
s
o
f
s
ize
3
×
3
.
R
e
L
U
i
s
also
u
s
e
d
R
eL
U
f
o
r
ac
tiv
atio
n
f
u
n
ctio
n
.
Nex
t
th
e
f
latten
ed
lay
e
r
.
T
h
i
s
lay
er
r
ef
er
s
to
th
e
p
r
o
ce
s
s
o
f
co
n
v
e
r
tin
g
a
m
u
ltid
im
en
s
io
n
al
ar
r
ay
(
e.
g
.
,
a
2
D)
in
to
a
o
n
e
-
d
im
en
s
io
n
al
ar
r
ay
.
Flatten
in
g
b
r
id
g
es
th
is
g
ap
b
y
r
esh
ap
in
g
th
e
2
D
f
e
atu
r
e
m
ap
s
in
to
a
1
D
ar
r
ay
,
wh
ich
ca
n
th
en
b
e
f
ed
in
to
th
e
d
en
s
e
lay
e
r
s
f
o
r
f
u
r
th
e
r
p
r
o
ce
s
s
in
g
an
d
d
ec
is
io
n
-
m
a
k
in
g
.
v)
T
h
e
d
r
o
p
o
u
t
lay
er
is
th
e
lay
er
b
ef
o
r
e
th
e
d
e
n
s
e
lay
er
.
Dr
o
p
o
u
t
is
a
f
o
r
m
o
f
r
eg
u
lar
izatio
n
th
at
h
elp
s
im
p
r
o
v
e
t
h
e
g
en
e
r
aliza
tio
n
o
f
a
n
eu
r
al
n
etwo
r
k
b
y
r
ed
u
c
in
g
o
v
e
r
f
itti
n
g
.
Ov
e
r
f
itti
n
g
o
cc
u
r
s
wh
en
a
m
o
d
el
lear
n
s
to
p
er
f
o
r
m
well
o
n
th
e
tr
ain
in
g
d
ata
b
u
t
p
er
f
o
r
m
s
p
o
o
r
ly
o
n
u
n
s
ee
n
d
ata.
T
h
e
n
u
m
b
er
o
f
0
.
2
o
r
2
0
%
r
ef
er
s
to
d
u
r
i
n
g
tr
ain
in
g
,
wh
ich
m
ea
n
s
,
2
0
%
o
f
th
e
n
eu
r
o
n
s
in
th
e
f
ir
s
t
d
en
s
e
lay
er
w
ill
b
e
r
an
d
o
m
l
y
d
r
o
p
p
ed
,
h
elp
in
g
p
r
ev
en
t
o
v
er
f
itti
n
g
an
d
im
p
r
o
v
i
n
g
th
e
m
o
d
el'
s
ab
ilit
y
to
g
en
e
r
alize
to
n
e
w
d
ata.
v
i)
L
astl
y
,
th
e
d
en
s
e
lay
er
is
als
o
k
n
o
wn
as
a
f
u
lly
co
n
n
ec
ted
lay
er
b
ec
au
s
e
ea
ch
n
eu
r
o
n
(
o
r
u
n
it)
in
th
e
d
en
s
e
lay
er
is
co
n
n
ec
te
d
to
ev
er
y
n
eu
r
o
n
i
n
th
e
p
r
ev
io
u
s
lay
er
an
d
e
v
er
y
n
eu
r
o
n
in
th
e
s
u
b
s
eq
u
en
t
lay
er
.
T
h
e
n
u
m
b
e
r
1
2
8
in
th
e
d
en
s
e
lay
er
r
ef
er
s
to
th
is
lay
er
h
a
s
1
2
8
n
eu
r
o
n
s
with
a
R
eL
U
f
o
r
ac
tiv
atio
n
f
u
n
ctio
n
.
Fig
u
r
e
2
.
T
h
e
p
r
o
p
o
s
ed
a
r
ch
it
ec
tu
r
e
o
f
d
ee
p
lear
n
in
g
2
.
4
.
P
er
f
o
r
m
a
nce
ev
a
lua
t
io
n
T
o
ass
ess
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
p
r
o
p
o
s
ed
class
if
icatio
n
m
e
th
o
d
o
lo
g
y
,
we
u
tili
ze
ac
c
u
r
ac
y
an
aly
s
is
f
o
r
tr
ain
i
n
g
,
v
alid
atio
n
,
an
d
te
s
tin
g
.
Ad
d
itio
n
ally
,
we
in
co
r
p
o
r
ate
an
an
aly
s
is
o
f
th
e
lo
s
s
f
u
n
ctio
n
to
d
eter
m
in
e
wh
eth
er
th
e
m
o
d
el
is
o
v
er
f
itti
n
g
o
r
n
o
t.
I
n
(
2
)
an
d
(
3
)
ar
e
e
m
p
lo
y
ed
to
ca
lcu
late
b
o
th
ac
c
u
r
ac
y
an
d
lo
s
s
.
=
×
1
0
0
%
(
2
)
(
−
)
=
−
1
∑
(
)
=
1
(
3
)
W
h
er
e
N
is
th
e
n
u
m
b
er
o
f
s
am
p
les an
d
p
i
,
y
i
is
th
e
p
r
e
d
icted
p
r
o
b
a
b
ilit
y
o
f
th
e
tr
u
e
class
lab
el
y
i
f
o
r
s
am
p
le
i.
3.
RE
SU
L
T
S
AND
D
I
S
CU
SS
I
O
N
I
n
th
is
s
tu
d
y
,
an
ass
ess
m
en
t
o
f
th
e
ap
p
r
o
p
r
iaten
ess
o
f
s
tate
-
of
-
th
e
-
a
r
t
d
ee
p
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
f
o
r
th
e
task
o
f
im
a
g
e
class
if
icatio
n
was
d
o
n
e.
T
h
i
s
f
o
cu
se
s
o
n
tu
n
in
g
h
y
p
er
p
ar
a
m
eter
s
s
u
ch
as
th
e
n
u
m
b
er
o
f
b
atch
s
ize
an
d
n
u
m
b
er
o
f
ep
o
ch
s
.
T
h
e
r
esu
lts
o
f
th
e
ex
p
er
im
en
t
a
r
e
d
is
cu
s
s
e
d
in
Fig
u
r
e
s
3
-
5
an
d
T
ab
le
s
1
-
3.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
14
,
No
.
2
,
J
u
n
e
2
0
2
5
:
3
8
4
-
393
388
3
.
1
.
H
a
rdwa
re
a
nd
s
o
f
t
wa
re
T
h
e
b
aselin
e
s
y
s
tem
u
s
ed
in
o
u
r
ev
alu
atio
n
is
a
PC
.
P
C
c
o
n
f
ig
u
r
atio
n
f
o
r
th
e
p
r
o
ject,
wh
er
e
th
e
p
r
o
ce
s
s
o
r
is
an
1
1
t
h
Gen
er
ati
o
n
I
n
tel(
R
)
,
C
o
r
e
(
T
M)
i5
-
1
1
3
5
G7
@
2
.
4
0
GHz
,
2
4
1
9
M
h
z,
4
C
o
r
es
.
OS
is
Mi
cr
o
s
o
f
t Wi
n
d
o
ws 1
0
,
Ph
y
to
n
,
Ker
as
,
an
d
J
u
p
y
ter
L
a
b
ar
e
u
s
ed
f
o
r
im
p
lem
en
tatio
n
[
3
8
]
,
[
3
9
]
.
3
.
2
.
M
o
del a
cc
ura
cy
a
nd
m
o
del lo
s
s
I
n
d
ee
p
lear
n
i
n
g
m
o
d
el
ac
cu
r
ac
y
an
d
m
o
d
el
lo
s
s
ar
e
p
lo
t
ted
to
s
h
o
w
th
e
ac
cu
r
ac
y
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el.
I
n
m
o
d
el
ac
cu
r
ac
y
,
t
r
ain
(
tr
ain
in
g
)
,
a
n
d
v
al
(
v
ali
d
ate)
ac
cu
r
ac
y
v
er
s
e
e
p
o
ch
i
s
p
lo
tted
.
T
r
ain
in
g
ac
cu
r
ac
y
r
e
p
r
esen
t
s
th
e
p
er
ce
n
tag
e
o
f
c
o
r
r
ec
tly
class
if
ied
e
x
am
p
les
in
th
e
tr
a
in
i
n
g
d
atase
t
an
d
th
e
aim
is
to
m
ea
s
u
r
e
h
o
w
well
th
e
m
o
d
el
f
its
th
e
tr
ain
in
g
d
ata
d
u
r
in
g
t
h
e
tr
ain
in
g
p
r
o
ce
s
s
.
V
alid
ate
ac
cu
r
ac
y
s
h
o
ws
th
e
p
er
ce
n
tag
e
o
f
co
r
r
ec
tly
class
if
ied
ex
am
p
les
in
th
e
v
alid
atio
n
d
ataset,
wh
ich
t
h
e
m
o
d
el
h
as
n
o
t
s
ee
n
d
u
r
in
g
tr
ain
in
g
.
T
h
e
h
ig
h
e
r
tr
ai
n
in
g
ac
cu
r
ac
y
an
d
v
alid
atio
n
ac
c
u
r
ac
y
r
e
p
r
esen
t
a
n
id
ea
l
s
ce
n
ar
i
o
wh
er
e
th
e
m
o
d
el
p
er
f
o
r
m
s
well
o
n
b
o
th
th
e
tr
ain
in
g
an
d
v
alid
atio
n
d
atasets
.
I
t
s
u
g
g
ests
th
at
th
e
m
o
d
el
h
as
lear
n
ed
th
e
u
n
d
er
ly
i
n
g
p
atter
n
s
in
th
e
d
ata
an
d
is
g
en
e
r
alize
d
ef
f
ec
tiv
ely
to
n
ew
ex
am
p
les.
Nex
t
,
th
e
m
o
d
el
lo
s
s
,
tr
ain
in
g
,
an
d
v
alid
ate
lo
s
s
v
er
s
e
e
p
o
ch
ar
e
also
p
lo
tted
.
T
r
ai
n
in
g
lo
s
s
r
ef
er
s
to
th
e
er
r
o
r
o
r
d
is
cr
ep
an
cy
b
etwe
en
th
e
p
r
ed
icted
o
u
tp
u
ts
o
f
a
d
ee
p
n
eu
r
al
n
etwo
r
k
an
d
th
e
ac
tu
al
tar
g
et
o
u
tp
u
ts
d
u
r
in
g
th
e
tr
ain
in
g
p
r
o
ce
s
s
.
T
h
e
v
alid
atio
n
lo
s
s
s
er
v
es
as
a
p
r
o
x
y
f
o
r
h
o
w
well
th
e
m
o
d
el
is
g
en
er
alizin
g
to
n
ew
d
ata.
A
lo
wer
v
alid
atio
n
lo
s
s
in
d
icate
s
th
at
th
e
m
o
d
el
is
m
ak
in
g
m
o
r
e
ac
c
u
r
ate
p
r
e
d
ictio
n
s
o
n
u
n
s
ee
n
ex
am
p
les
f
r
o
m
t
h
e
v
alid
ati
o
n
d
ataset.
T
h
e
m
o
d
el
lo
s
s
aim
s
to
ac
h
iev
e
lo
w
t
r
ain
in
g
an
d
v
alid
atio
n
lo
s
s
es
s
im
u
ltan
eo
u
s
ly
,
in
d
icatin
g
th
at
th
e
m
o
d
el
h
as
lear
n
ed
m
e
an
in
g
f
u
l
p
atter
n
s
f
r
o
m
th
e
tr
ain
in
g
d
ata
an
d
ca
n
g
en
er
alize
ef
f
ec
tiv
el
y
to
n
ew,
u
n
s
ee
n
d
ata.
Fig
u
r
e
3
s
h
o
w
s
th
e
m
o
d
el
o
f
m
o
d
el
ac
cu
r
ac
y
(
Fig
u
r
e
3
(
a)
)
an
d
m
o
d
el
lo
s
s
(
Fig
u
r
e
3
(
b
)
)
a
cc
o
r
d
in
g
to
tr
ain
an
d
v
alid
atio
n
d
ata
ac
c
o
r
d
in
g
to
b
atc
h
s
ize=
1
6
,
a
n
u
m
b
er
o
f
e
p
o
ch
s
0
to
2
5
,
a
n
d
th
e
o
p
tim
izatio
n
alg
o
r
ith
m
is
Ad
a
m
.
I
n
Fig
u
r
e
3
,
wh
en
th
e
ep
o
c
h
v
alu
e
ap
p
r
o
ac
h
es
1
5
to
2
5
,
th
e
m
o
d
el
ac
cu
r
ac
y
(
Fig
u
r
e
3
(
a)
)
s
h
o
ws
a
c
o
n
s
is
ten
t
in
cr
ea
s
e
in
ac
cu
r
ac
y
,
r
ea
ch
i
n
g
a
p
latea
u
wh
er
e
it
r
em
ai
n
s
s
tab
le.
Similar
ly
,
in
th
e
m
o
d
el
lo
s
s
(
Fig
u
r
e
3
(
b
)
)
,
th
e
tr
ai
n
in
g
lo
s
s
d
ec
r
ea
s
es st
ea
d
ily
an
d
th
en
s
tab
ilizes with
in
th
e
s
am
e
ep
o
ch
r
a
n
g
e.
Fig
u
r
e
4
s
h
o
w
s
th
e
m
o
d
el
o
f
m
o
d
el
ac
cu
r
ac
y
(
Fig
u
r
e
4
(
a)
)
an
d
m
o
d
el
lo
s
s
ac
co
r
d
in
g
(
Fig
u
r
e
4
(
b
)
)
to
tr
ain
an
d
v
alid
atio
n
d
ata
ac
c
o
r
d
in
g
to
b
atc
h
s
ize=
3
2
,
a
n
u
m
b
er
o
f
e
p
o
ch
s
0
to
2
5
,
a
n
d
th
e
o
p
tim
izatio
n
alg
o
r
ith
m
a
r
e
Ad
am
.
I
n
Fig
u
r
e
4
,
th
e
p
atter
n
clo
s
ely
m
ir
r
o
r
s
th
at
o
f
Fig
u
r
e
3
.
W
h
en
t
h
e
e
p
o
ch
v
alu
e
r
ea
c
h
es
1
5
to
2
5
,
th
e
m
o
d
el
ac
c
u
r
ac
y
(
Fig
u
r
e
4
(
a)
)
e
x
h
ib
its
a
co
n
s
is
ten
t
in
cr
ea
s
e
an
d
th
en
s
tab
il
izes.
Similar
ly
,
th
e
m
o
d
el
lo
s
s
(
Fig
u
r
e
4
(
b
)
)
d
em
o
n
s
tr
ates
a
d
ec
r
ea
s
in
g
tr
en
d
an
d
ev
e
n
tu
ally
s
tab
ilizes
with
in
th
e
s
am
e
ep
o
c
h
r
an
g
e.
Fig
u
r
e
5
s
h
o
w
s
th
e
m
o
d
el
o
f
m
o
d
el
ac
cu
r
ac
y
(
Fig
u
r
e
5
(
a)
)
an
d
m
o
d
el
lo
s
s
ac
co
r
d
in
g
(
Fig
u
r
e
5
(
b
)
)
to
tr
ain
an
d
v
alid
atio
n
d
ata
ac
c
o
r
d
in
g
to
b
atc
h
s
ize=
6
4
,
a
n
u
m
b
er
o
f
e
p
o
ch
s
0
to
2
5
,
a
n
d
th
e
o
p
tim
izatio
n
alg
o
r
ith
m
ar
e
Ad
am
.
I
n
Fig
u
r
e
5
,
th
e
p
atter
n
clo
s
ely
r
esem
b
les
th
at
o
f
Fig
u
r
es
3
an
d
4
.
As
th
e
ep
o
ch
v
alu
e
n
ea
r
s
1
5
to
2
5
,
th
e
ac
c
u
r
ac
y
m
o
d
el
(
Fig
u
r
e
5
(
a
)
)
d
is
p
lay
s
a
s
tead
y
in
cr
ea
s
e
f
o
llo
wed
b
y
s
tab
ilit
y
.
Similar
ly
,
in
th
e
m
o
d
el
lo
s
s
(
tr
ain
in
g
lo
s
s
)
(
Fig
u
r
e
5
(
b
)
)
,
t
h
e
r
e'
s
a
d
ec
r
ea
s
in
g
tr
en
d
,
e
v
en
tu
ally
s
tab
ilizin
g
b
etwe
en
th
e
1
5
an
d
2
5
ep
o
c
h
s
.
(
a)
(
b
)
Fig
u
r
e
3
.
Mo
d
el
o
f
(
a)
m
o
d
el
ac
cu
r
ac
y
an
d
(
b
)
m
o
d
el
lo
s
s
ac
co
r
d
in
g
to
tr
ain
a
n
d
v
alid
atio
n
d
ata
ac
co
r
d
in
g
t
o
b
atch
s
ize=
1
6
,
n
u
m
b
e
r
o
f
e
p
o
ch
s
0
to
2
5
,
an
d
th
e
o
p
tim
izatio
n
alg
o
r
ith
m
is
Ad
am
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
C
o
mp
a
r
a
tive
s
tu
d
y
o
n
fin
e
-
t
u
n
in
g
d
ee
p
lea
r
n
in
g
mo
d
els fo
r
fr
u
it a
n
d
ve
g
eta
b
le
…
(
A
b
d
R
a
s
id
Ma
ma
t)
389
(
a)
(
b
)
Fig
u
r
e
4
.
Mo
d
el
o
f
(
a)
m
o
d
el
ac
cu
r
ac
y
an
d
(
b
)
m
o
d
el
lo
s
s
ac
co
r
d
in
g
to
tr
ain
a
n
d
v
alid
atio
n
d
ata
ac
co
r
d
in
g
t
o
b
atch
s
ize=
3
2
,
n
u
m
b
e
r
o
f
e
p
o
ch
s
0
to
2
5
,
an
d
th
e
o
p
tim
izatio
n
alg
o
r
ith
m
is
Ad
am
(
a)
(
b
)
Fig
u
r
e
5
.
Mo
d
el
o
f
(
a)
m
o
d
el
ac
cu
r
ac
y
an
d
(
b
)
m
o
d
el
lo
s
s
ac
co
r
d
in
g
to
tr
ain
a
n
d
v
alid
atio
n
d
ata
ac
co
r
d
in
g
t
o
b
atch
s
ize=
6
4
,
n
u
m
b
e
r
o
f
e
p
o
ch
s
0
to
2
5
,
an
d
th
e
o
p
tim
izatio
n
alg
o
r
ith
m
is
Ad
am
3
.
3
.
Co
m
pa
riso
n o
f
a
cc
ura
c
y
ba
s
ed
o
n diff
er
ent
ba
t
ch
s
izes a
t
a
nu
m
ber
o
f
epo
chs
1
0
B
ased
o
n
T
a
b
le
1
,
in
cr
ea
s
in
g
th
e
b
atch
s
ize
f
r
o
m
1
6
to
6
4
r
esu
lted
in
a
n
in
c
r
ea
s
e
in
ac
cu
r
ac
y
(
b
o
th
tr
ain
in
g
an
d
v
alid
atio
n
)
o
f
2
.
9
0
an
d
0
.
8
6
%,
r
esp
ec
tiv
el
y
.
T
h
is
d
em
o
n
s
tr
ates
th
at
th
e
m
o
d
el'
s
ac
cu
r
ac
y
is
d
ir
ec
tly
p
r
o
p
o
r
tio
n
al
to
t
h
e
in
cr
ea
s
e
in
b
atch
s
ize
an
d
in
v
er
s
ely
p
r
o
p
o
r
tio
n
al
to
th
e
tr
ain
in
g
lo
s
s
.
C
o
n
s
eq
u
en
tly
,
th
e
m
o
d
el
b
ec
o
m
es
m
o
r
e
ac
c
u
r
ate
as
th
e
b
a
tch
s
ize
in
cr
ea
s
es.
Ho
wev
er
,
i
t's
n
o
tab
le
th
at
th
e
av
er
ag
e
tim
e
tak
e
n
p
er
s
tep
o
r
b
atch
d
u
r
in
g
tr
ai
n
in
g
(
m
s
/s
tep
)
also
in
cr
ea
s
es wi
th
th
e
b
atch
s
ize.
B
ased
o
n
T
ab
le
2
,
th
e
s
am
e
p
atter
n
as
T
ab
le
1
is
o
b
s
er
v
e
d
.
I
n
cr
ea
s
in
g
th
e
b
atch
s
ize
f
r
o
m
1
6
t
o
6
4
r
esu
lted
in
an
i
n
cr
ea
s
e
in
ac
cu
r
ac
y
(
b
o
t
h
tr
ain
i
n
g
an
d
v
a
lid
atio
n
)
o
f
1
.
7
4
a
n
d
0
.
5
7
%,
r
esp
ec
tiv
ely
.
T
h
is
in
cr
ea
s
e
in
ac
c
u
r
ac
y
in
d
icate
s
a
m
o
r
e
ac
c
u
r
ate
m
o
d
el
with
a
lar
g
er
b
atch
s
ize.
Ad
d
itio
n
a
lly
,
it'
s
n
o
tewo
r
th
y
th
at
th
e
av
er
ag
e
tim
e
tak
en
p
e
r
s
tep
o
r
b
atc
h
d
u
r
in
g
tr
ai
n
in
g
d
ec
r
ea
s
ed
f
r
o
m
b
atch
s
ize
1
6
t
o
6
4
.
I
n
T
ab
le
3
,
it
also
s
h
o
ws
th
e
s
am
e
p
atter
n
as
T
ab
les
1
an
d
2
.
T
h
e
ac
cu
r
ac
y
v
alu
e
i
n
cr
ea
s
es
(
b
o
th
tr
ain
in
g
an
d
v
alid
atio
n
)
b
y
0
.
9
3
an
d
0
.
8
6
%,
wh
en
th
e
b
atch
s
ize
in
cr
ea
s
es
f
r
o
m
1
6
to
6
4
.
Me
an
wh
ile,
th
e
av
er
ag
e
tim
e
tak
e
n
p
er
s
tep
o
r
b
atch
d
u
r
in
g
tr
ai
n
in
g
r
e
d
u
ce
d
f
r
o
m
b
atch
s
ize
1
6
to
6
4
.
B
ased
o
n
th
e
r
esu
lts
,
th
e
ac
cu
r
ac
y
im
p
r
o
v
es
with
a
n
in
c
r
ea
s
e
in
b
atch
s
ize
(
r
esu
ltin
g
in
m
o
r
e
d
ata)
an
d
th
e
n
u
m
b
e
r
o
f
ep
o
c
h
s
(
al
lo
win
g
th
e
m
o
d
el
m
o
r
e
o
p
p
o
r
tu
n
ities
to
lear
n
f
r
o
m
th
e
d
ata
,
lead
in
g
to
b
etter
co
n
v
er
g
en
ce
an
d
r
e
d
u
ce
d
u
n
d
er
f
itti
n
g
)
.
T
h
e
r
esu
lts
illu
s
tr
ate
th
e
h
ig
h
est
ac
c
u
r
ac
y
o
b
tain
ed
,
wh
en
t
u
n
in
g
p
ar
am
eter
s
f
o
r
b
atc
h
s
ize
is
6
4
an
d
th
e
n
u
m
b
e
r
o
f
ep
o
ch
s
is
2
5
(
T
ab
le
3
)
.
T
h
is
is
b
ec
au
s
e
a
lar
g
er
b
atch
s
ize
an
d
m
o
r
e
ep
o
ch
s
p
r
o
v
id
e
a
m
o
r
e
s
tab
le
an
d
ac
cu
r
ate
esti
m
ate
o
f
th
e
g
r
ad
ie
n
t
o
f
th
e
lo
s
s
f
u
n
ctio
n
,
r
e
d
u
cin
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
14
,
No
.
2
,
J
u
n
e
2
0
2
5
:
3
8
4
-
393
390
n
o
is
e
an
d
f
lu
ct
u
atio
n
s
in
th
e
lear
n
in
g
p
r
o
ce
s
s
,
wh
ich
u
ltim
ately
lead
s
to
a
m
o
r
e
ac
c
u
r
at
e
p
r
o
p
o
s
ed
m
o
d
el.
T
h
er
ef
o
r
e,
it
ca
n
p
r
o
d
u
ce
b
ett
er
g
e
n
er
aliza
tio
n
to
u
n
s
ee
n
d
a
ta,
as
th
e
m
o
d
el
ca
n
av
o
id
o
v
er
f
itti
n
g
b
y
t
r
ain
in
g
o
n
m
o
r
e
d
ata
p
o
in
ts
in
e
ac
h
iter
atio
n
.
T
h
is
ap
p
r
o
ac
h
c
an
en
h
an
ce
t
h
e
ac
cu
r
ac
y
o
f
v
alid
atio
n
r
esu
lts
.
Ad
d
itio
n
ally
,
lar
g
er
b
atch
s
ize
s
an
d
m
o
r
e
ep
o
ch
s
b
e
n
ef
it
m
o
d
er
n
g
r
ap
h
ics
p
r
o
ce
s
s
in
g
u
n
its
(
GPUs
)
o
r
ten
s
o
r
p
r
o
ce
s
s
in
g
u
n
its
(
T
PUs
)
,
allo
win
g
f
o
r
f
aster
tr
ain
in
g
wh
ile
m
ain
tain
in
g
o
r
ev
en
im
p
r
o
v
in
g
m
o
d
el
ac
c
u
r
ac
y
.
T
ab
le
1
.
C
o
m
p
a
r
is
o
n
o
f
ac
cu
r
ac
y
an
d
l
o
s
s
u
s
in
g
d
if
f
er
e
n
t b
a
tch
s
izes f
o
r
a
n
u
m
b
er
o
f
ep
o
c
h
s
=1
0
B
a
t
c
h
S
i
z
e
N
u
m
o
f
e
p
o
c
h
s
Tr
a
i
n
i
n
g
a
c
c
u
r
a
c
y
(
%)
Tr
a
i
n
i
n
g
l
o
ss
(
%)
V
a
l
i
d
a
t
i
o
n
a
c
c
u
r
a
c
y
(
%)
V
a
l
i
d
a
t
i
o
n
l
o
ss
(
%)
A
v
e
r
a
g
e
t
i
me
t
a
k
e
n
p
e
r
st
e
p
o
r
b
a
t
c
h
d
u
r
i
n
g
t
r
a
i
n
i
n
g
(
ms/s
t
e
p
)
16
10
0
.
9
4
6
6
0
.
2
0
8
7
0
.
9
4
8
7
0
.
4
8
3
3
2
1
6
32
10
0
.
9
5
9
3
0
.
1
7
6
8
0
.
9
5
1
6
0
.
4
1
2
7
3
9
9
64
10
0
.
9
7
4
0
0
.
1
1
9
2
0
.
9
5
7
3
0
.
2
6
0
7
7
3
1
T
ab
le
2
.
C
o
m
p
a
r
is
o
n
o
f
ac
cu
r
ac
y
b
ased
o
n
d
if
f
e
r
en
t b
atc
h
s
izes f
o
r
a
n
u
m
b
er
o
f
e
p
o
ch
s
=1
5
B
a
t
c
h
S
i
z
e
N
u
m
o
f
Ep
o
c
h
s
Tr
a
i
n
i
n
g
a
c
c
u
r
a
c
y
(
%)
Tr
a
i
n
i
n
g
l
o
ss
(
%)
V
a
l
i
d
a
t
i
o
n
a
c
c
u
r
a
c
y
(
%)
V
a
l
i
d
a
t
i
o
n
l
o
ss
(
%)
A
v
e
r
a
g
e
t
i
me
t
a
k
e
n
p
e
r
st
e
p
o
r
b
a
t
c
h
d
u
r
i
n
g
t
r
a
i
n
i
n
g
(
ms/s
t
e
p
)
16
15
0
.
9
6
9
3
0
.
1
2
6
2
0
.
9
4
8
7
0
.
5
3
9
1
2
1
2
32
15
0
.
9
8
3
7
0
.
0
8
4
0
0
.
9
4
0
2
0
.
4
5
8
7
3
9
6
64
15
0
.
9
8
6
1
0
.
0
6
7
9
0
.
9
5
4
4
0
.
3
0
9
1
7
4
3
T
ab
le
3
.
C
o
m
p
a
r
is
o
n
o
f
ac
cu
r
ac
y
b
ased
o
n
d
if
f
e
r
en
t b
atc
h
s
izes f
o
r
a
n
u
m
b
er
o
f
e
p
o
ch
s
=2
5
B
a
t
c
h
S
i
z
e
N
u
m
o
f
Ep
o
c
h
s
Tr
a
i
n
i
n
g
a
c
c
u
r
a
c
y
(
%)
Tr
a
i
n
i
n
g
l
o
ss
(
%)
V
a
l
i
d
a
t
i
o
n
a
c
c
u
r
a
c
y
(
%)
V
a
l
i
d
a
t
i
o
n
l
o
ss
(
%)
A
v
e
r
a
g
e
t
i
me
t
a
k
e
n
p
e
r
st
e
p
o
r
b
a
t
c
h
d
u
r
i
n
g
t
r
a
i
n
i
n
g
(
ms/s
t
e
p
)
16
25
0
.
9
8
1
1
0
.
0
6
7
5
0
.
9
5
4
4
0
.
4
7
5
2
2
1
6
32
25
0
.
9
8
4
5
0
.
0
6
5
5
0
.
9
5
4
8
0
.
4
1
7
3
4
0
1
64
25
0
.
9
9
0
2
0
.
0
6
0
6
0
.
9
6
3
0
0
.
2
8
6
9
6
3
6
3
.
4
.
T
esting
da
t
a
f
o
r
t
he
pro
po
s
ed
m
o
del
B
ased
o
n
th
e
r
esu
lts
f
r
o
m
s
u
b
s
ec
tio
n
3
.
3
,
th
e
p
r
o
p
o
s
ed
m
o
d
el
was
tr
ain
ed
u
s
in
g
2
5
e
p
o
ch
s
an
d
b
atc
h
s
izes
o
f
1
6
,
3
2
,
an
d
6
4
.
T
h
e
ac
cu
r
ac
y
o
f
th
e
tr
ain
in
g
d
at
aset
f
o
r
th
e
p
r
o
p
o
s
ed
m
o
d
el
is
p
r
esen
ted
,
with
ex
am
p
les
f
r
o
m
th
e
ap
p
le
an
d
ca
b
b
ag
e
class
es
an
d
th
e
r
esu
lt
d
is
p
lay
ed
in
T
a
b
le
4
.
Acc
o
r
d
in
g
to
th
e
e
x
am
p
le
o
f
th
is
clas
s
,
it
wa
s
f
o
u
n
d
th
at
th
e
av
er
ag
e
ac
cu
r
ac
y
o
f
t
h
e
tr
ain
in
g
d
ataset
in
cr
ea
s
es
wh
en
th
e
b
atch
s
ize
an
d
n
u
m
b
er
o
f
ep
o
ch
s
in
cr
ea
s
e.
I
t sh
o
ws
in
cr
ea
s
in
g
th
e
b
atch
s
iz
e
an
d
n
u
m
b
e
r
o
f
ep
o
c
h
s
lead
s
to
h
ig
h
er
ac
cu
r
ac
y
o
n
th
e
test
in
g
d
ataset,
it
in
d
ic
ates
th
at
th
e
ch
o
s
en
b
atch
s
iz
e
an
d
n
u
m
b
e
r
o
f
ep
o
c
h
s
ar
e
co
n
d
u
ci
v
e
to
b
etter
g
en
er
aliza
tio
n
a
n
d
o
v
er
all
m
o
d
el
p
er
f
o
r
m
an
ce
.
T
ab
le
4
.
Av
e
r
ag
e
ac
cu
r
ac
y
o
f
th
e
test
in
g
d
ata
f
o
r
class
ap
p
le
an
d
wate
r
m
elo
n
A
v
e
r
a
g
e
a
c
c
u
r
a
c
y
o
f
t
r
a
i
n
i
n
g
(
%)
o
f
2
5
n
u
m
b
e
r
s o
f
e
p
o
c
h
s
C
l
a
s
s
B
a
t
c
h
si
z
e
A
p
p
l
e
W
a
t
e
r
m
e
l
o
n
16
9
3
.
8
3
0
0
8
9
.
8
8
8
3
32
9
6
.
2
5
5
0
9
7
.
0
5
0
0
64
9
9
.
5
7
0
0
9
9
.
7
6
0
0
4.
CO
NCLU
SI
O
N
T
h
is
wo
r
k
h
as
d
ev
elo
p
ed
to
m
o
d
el
f
r
u
it
an
d
v
eg
etab
le
s
u
s
in
g
f
in
e
-
tu
n
in
g
h
y
p
er
p
ar
am
ete
r
s
o
f
C
NN.
Fro
m
th
e
ex
p
er
im
e
n
ts
,
a
p
r
o
p
o
s
ed
m
o
d
el
with
a
co
m
b
in
atio
n
o
f
th
e
b
atch
s
ize=
6
4
a
n
d
th
e
n
u
m
b
er
o
f
ep
o
ch
s
=2
5
,
p
r
o
d
u
ce
s
th
e
m
o
s
t
o
p
tim
al
m
o
d
el
with
an
a
cc
u
r
ac
y
v
al
u
e
(
tr
ain
i
n
g
)
o
f
9
9
.
0
2
%,
w
h
ile
th
e
v
alid
atio
n
is
9
5
.
7
3
%
an
d
th
e
l
o
s
s
is
6
.
0
6
%
(
m
in
im
u
m
)
.
I
t
ca
n
b
e
co
n
clu
d
e
d
th
at
an
i
n
cr
ea
s
e
in
b
atch
s
ize
an
d
n
u
m
b
er
o
f
ep
o
c
h
s
is
d
ir
ec
tly
p
r
o
p
o
r
tio
n
al
to
ac
cu
r
ac
y
f
o
r
b
o
th
tr
ai
n
in
g
a
n
d
v
alid
a
tio
n
an
d
in
v
er
s
ely
p
r
o
p
o
r
tio
n
al
to
tr
ain
in
g
an
d
v
alid
atio
n
lo
s
s
an
d
th
is
r
elatio
n
s
h
ip
lead
s
to
m
o
r
e
ac
cu
r
ate
m
o
d
el
p
r
ed
ictio
n
s
.
C
o
n
s
id
er
in
g
th
is
r
esu
lt,
wh
ich
in
v
o
lv
es th
e
r
elatio
n
s
h
ip
b
etw
ee
n
b
atch
s
ize
an
d
th
e
n
u
m
b
er
o
f
ep
o
ch
s
,
th
er
e
is
an
o
p
p
o
r
tu
n
ity
f
o
r
r
esear
ch
e
r
s
to
ex
p
lo
r
e
f
u
r
th
e
r
s
tu
d
ies
o
n
th
e
f
in
e
-
tu
n
in
g
r
elatio
n
s
h
ip
s
b
etwe
en
o
th
er
h
y
p
er
p
a
r
am
eter
s
to
p
r
o
d
u
ce
a
m
o
r
e
ac
cu
r
ate
p
r
ed
ictio
n
m
o
d
el
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
p
r
o
ject
is
f
u
n
d
ed
b
y
th
e
C
en
ter
f
o
r
R
esear
ch
E
x
ce
llen
ce
,
I
n
cu
b
atio
n
Ma
n
ag
e
m
e
n
t
C
en
ter
,
Un
iv
er
s
ity
Su
ltan
Z
ain
al
Ab
id
in
,
Ma
lay
s
ia.
Un
d
er
Pro
ject
C
o
d
e
is
Un
iSZA/2
0
2
3
/DPU 1
.
0
/0
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
C
o
mp
a
r
a
tive
s
tu
d
y
o
n
fin
e
-
t
u
n
in
g
d
ee
p
lea
r
n
in
g
mo
d
els fo
r
fr
u
it a
n
d
ve
g
eta
b
le
…
(
A
b
d
R
a
s
id
Ma
ma
t)
391
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Ab
d
R
asid
Ma
m
at
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Mo
h
am
ad
Af
e
n
d
ee
Mo
h
am
ed
✓
✓
✓
✓
✓
✓
Mo
h
d
Fad
zil
Ab
d
Kad
ir
✓
✓
✓
✓
✓
✓
No
r
k
h
air
an
i
Ab
d
u
l
R
awi
✓
✓
✓
✓
Azim
Z
alih
a
Ab
d
Aziz
✓
✓
✓
✓
✓
W
an
Su
r
y
an
i Wan
Awa
n
g
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
T
h
e
au
th
o
r
s
d
ec
lar
e
th
at
th
e
r
e
ar
e
n
o
c
o
n
f
licts
o
f
in
ter
est r
e
g
ar
d
in
g
th
e
p
u
b
licatio
n
o
f
th
is
p
ap
er
.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ataset
u
s
ed
to
wo
r
k
b
e
f
r
u
it
_
v
eg
eta
b
le’
s
d
ataset
an
d
wh
ich
is
p
u
b
licly
av
ailab
le
o
n
Kag
g
le
wh
ich
is
p
u
b
licly
av
ailab
le.
I
t
f
r
ee
a
n
d
d
o
wn
l
o
ad
f
r
o
m
h
ttp
s
://www.
k
ag
g
le.
co
m
/d
atasets
/
k
r
itik
s
eth
/f
r
u
it
-
an
d
-
v
eg
etab
le
-
im
ag
e
-
r
ec
o
g
n
itio
n
.
RE
F
E
R
E
NC
E
S
[
1
]
M
.
W
u
,
J.
Z
h
o
u
,
Y
.
P
e
n
g
,
S
.
W
a
n
g
,
a
n
d
Y
.
Zh
a
n
g
,
“
D
e
e
p
l
e
a
r
n
i
n
g
f
o
r
i
m
a
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
:
a
r
e
v
i
e
w
,
”
i
n
P
ro
c
e
e
d
i
n
g
s
o
f
2
0
2
3
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
M
e
d
i
c
a
l
I
m
a
g
i
n
g
a
n
d
C
o
m
p
u
t
e
r
-
Ai
d
e
d
D
i
a
g
n
o
si
s
(
MIC
AD
2
0
2
3
)
,
S
p
r
i
n
g
e
r
,
S
i
n
g
a
p
o
r
e
,
p
p
.
3
5
2
–
362
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
97
-
1
3
3
5
-
6
_
3
1
.
[
2
]
P
.
T.
Q
.
A
n
h
,
D
.
Q
.
T
h
u
y
e
t
,
a
n
d
Y
.
K
o
b
a
y
a
sh
i
,
“
I
mag
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
o
f
r
o
o
t
-
t
r
i
mm
e
d
g
a
r
l
i
c
u
s
i
n
g
m
u
l
t
i
-
l
a
b
e
l
a
n
d
m
u
l
t
i
-
c
l
a
ss
c
l
a
ss
i
f
i
c
a
t
i
o
n
w
i
t
h
d
e
e
p
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
Po
s
t
h
a
r
v
e
s
t
Bi
o
l
o
g
y
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
9
0
,
A
u
g
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
o
st
h
a
r
v
b
i
o
.
2
0
2
2
.
1
1
1
9
5
6
.
[
3
]
A
.
Z.
A
.
A
z
i
z
,
M
.
S
.
M
.
S
a
b
i
l
a
n
,
a
n
d
N
.
F
.
M
a
n
so
r
,
“
T
h
e
e
f
f
e
c
t
i
v
e
n
e
ss
o
f
u
si
n
g
d
e
p
t
h
i
m
a
g
e
s
f
o
r
g
a
i
t
-
b
a
se
d
g
e
n
d
e
r
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
I
AE
N
G
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
5
0
,
n
o
.
3
,
2
0
2
3
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
i
a
e
n
g
.
o
r
g
/
I
JC
S
/
i
ssu
e
s_
v
5
0
/
i
ssu
e
_
3
/
I
JC
S
_
5
0
_
3
_
0
3
.
p
d
f
[
4
]
A
.
C
a
r
b
o
n
e
,
“
C
a
n
c
e
r
c
l
a
ssi
f
i
c
a
t
i
o
n
a
t
t
h
e
c
r
o
ssr
o
a
d
s
,
”
C
a
n
c
e
rs
,
v
o
l
.
1
2
,
n
o
.
4
,
A
p
r
.
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
c
a
n
c
e
r
s
1
2
0
4
0
9
8
0
.
[
5
]
M
.
Z
u
b
a
i
r
,
S
.
W
a
n
g
,
a
n
d
N
.
A
l
i
,
“
A
d
v
a
n
c
e
d
a
p
p
r
o
a
c
h
e
s
t
o
b
r
e
a
st
c
a
n
c
e
r
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
n
d
d
i
a
g
n
o
si
s
,
”
Fr
o
n
t
i
e
rs
i
n
Ph
a
rm
a
c
o
l
o
g
y
,
v
o
l
.
1
1
,
F
e
b
.
2
0
2
1
,
d
o
i
:
1
0
.
3
3
8
9
/
f
p
h
a
r
.
2
0
2
0
.
6
3
2
0
7
9
.
[
6
]
C
.
C
.
U
k
w
u
o
ma,
Q
.
Z
h
i
g
u
a
n
g
,
M
.
B
.
B
i
n
H
e
y
a
t
,
L
.
A
l
i
,
Z.
A
l
m
a
sp
o
o
r
,
a
n
d
H
.
N
.
M
o
n
d
a
y
,
“
R
e
c
e
n
t
a
d
v
a
n
c
e
me
n
t
s
i
n
f
r
u
i
t
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s
,
”
M
a
t
h
e
m
a
t
i
c
a
l
Pr
o
b
l
e
m
s
i
n
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
2
0
2
2
,
p
p
.
1
–
2
9
,
Jan
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
2
/
9
2
1
0
9
4
7
.
[
7
]
R
.
B
u
t
u
n
e
r
,
I
.
C
i
n
a
r
,
Y
.
S
.
Ta
s
p
i
n
a
r
,
R
.
K
u
r
su
n
,
M
.
H
.
C
a
l
p
,
a
n
d
M
.
K
o
k
l
u
,
“
C
l
a
ssi
f
i
c
a
t
i
o
n
o
f
d
e
e
p
i
ma
g
e
f
e
a
t
u
r
e
s
o
f
l
e
n
t
i
l
v
a
r
i
e
t
i
e
s
w
i
t
h
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,”
Eu
r
o
p
e
a
n
F
o
o
d
R
e
se
a
rc
h
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
4
9
,
n
o
.
5
,
p
p
.
1
3
0
3
–
1
3
1
6
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
2
1
7
-
0
2
3
-
0
4
2
1
4
-
z.
[
8
]
N.
-
E.
-
A
.
M
i
mm
a
,
S
.
A
h
me
d
,
T.
R
a
h
ma
n
,
a
n
d
R
.
K
h
a
n
,
“
F
r
u
i
t
s
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
n
d
d
e
t
e
c
t
i
o
n
a
p
p
l
i
c
a
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
S
c
i
e
n
t
i
f
i
c
Pro
g
r
a
m
m
i
n
g
,
v
o
l
.
2
0
2
2
,
p
p
.
1
–
1
6
,
N
o
v
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
2
/
4
1
9
4
8
7
4
.
[
9
]
R
.
M
a
h
u
m
e
t
a
l
.
,
“
A
n
o
v
e
l
f
r
a
mew
o
r
k
f
o
r
p
o
t
a
t
o
l
e
a
f
d
i
s
e
a
se
d
e
t
e
c
t
i
o
n
u
si
n
g
a
n
e
f
f
i
c
i
e
n
t
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
,
”
H
u
m
a
n
a
n
d
Ec
o
l
o
g
i
c
a
l
Ri
s
k
Assessm
e
n
t
:
An
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
,
v
o
l
.
2
9
,
n
o
.
2
,
p
p
.
3
0
3
–
3
2
6
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
8
0
/
1
0
8
0
7
0
3
9
.
2
0
2
2
.
2
0
6
4
8
1
4
.
[
1
0
]
S
.
I
.
Ju
ma
,
Z.
M
u
d
a
,
M
.
A
.
M
o
h
a
m
e
d
,
a
n
d
W
.
M
o
h
a
me
d
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s f
o
r
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
s
y
st
e
m
:
a
r
e
v
i
e
w
,
”
J
o
u
rn
a
l
o
f
T
h
e
o
r
e
t
i
c
a
l
a
n
d
A
p
p
l
i
e
d
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
7
2
,
n
o
.
3
,
p
p
.
4
2
2
–
4
2
9
,
2
0
1
5
.
[
1
1
]
Z.
-
Q
.
Zh
a
o
,
P
.
Z
h
e
n
g
,
S
.
-
T
.
X
u
,
a
n
d
X
.
W
u
,
“
O
b
j
e
c
t
d
e
t
e
c
t
i
o
n
w
i
t
h
d
e
e
p
l
e
a
r
n
i
n
g
:
a
r
e
v
i
e
w
,
”
I
E
EE
T
ra
n
s
a
c
t
i
o
n
s
o
n
N
e
u
r
a
l
N
e
t
w
o
rks
a
n
d
L
e
a
rn
i
n
g
S
y
st
e
m
s
,
v
o
l
.
3
0
,
n
o
.
1
1
,
p
p
.
3
2
1
2
–
3
2
3
2
,
N
o
v
.
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
TN
N
LS
.
2
0
1
8
.
2
8
7
6
8
6
5
.
[
1
2
]
W
.
K
h
a
n
,
A
.
D
a
u
d
,
K
.
K
h
a
n
,
S
.
M
u
h
a
mm
a
d
,
a
n
d
R
.
H
a
q
,
“
E
x
p
l
o
r
i
n
g
t
h
e
f
r
o
n
t
i
e
r
s
o
f
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ss
i
n
g
:
a
c
o
m
p
r
e
h
e
n
si
v
e
o
v
e
r
v
i
e
w
o
f
k
e
y
c
h
a
l
l
e
n
g
e
s
a
n
d
e
m
e
r
g
i
n
g
t
r
e
n
d
s,
”
N
a
t
u
r
a
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ssi
n
g
J
o
u
r
n
a
l
,
v
o
l
.
4
,
S
e
p
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
l
p
.
2
0
2
3
.
1
0
0
0
2
6
.
[
1
3
]
H
.
H
e
r
man
,
A
.
S
u
s
a
n
t
o
,
T.
W
.
C
e
n
g
g
o
r
o
,
S
.
S
u
h
a
r
j
i
t
o
,
a
n
d
B
.
P
a
r
d
a
me
a
n
,
“
O
i
l
p
a
l
m
f
r
u
i
t
i
m
a
g
e
r
i
p
e
n
e
ss
c
l
a
ssi
f
i
c
a
t
i
o
n
w
i
t
h
c
o
m
p
u
t
e
r
v
i
si
o
n
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
v
i
s
u
a
l
a
t
t
e
n
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
T
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
,
El
e
c
t
ro
n
i
c
a
n
d
C
o
m
p
u
t
e
r
E
n
g
i
n
e
e
ri
n
g
(
J
T
EC
)
,
v
o
l
.
1
2
,
n
o
.
2
,
p
p
.
2
1
–
2
7
,
2
0
2
0
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
j
t
e
c
.
u
t
e
m.e
d
u
.
my
/
j
t
e
c
/
a
r
t
i
c
l
e
/
v
i
e
w
/
5
5
4
3
[
1
4
]
S
u
h
a
r
j
i
t
o
,
G
.
N
.
E
l
w
i
r
e
h
a
r
d
j
a
,
a
n
d
J.
S
.
P
r
a
y
o
g
a
,
“
O
i
l
p
a
l
m
f
r
e
sh
f
r
u
i
t
b
u
n
c
h
r
i
p
e
n
e
ss
c
l
a
ss
i
f
i
c
a
t
i
o
n
o
n
mo
b
i
l
e
d
e
v
i
c
e
s u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s,
”
C
o
m
p
u
t
e
rs
a
n
d
El
e
c
t
r
o
n
i
c
s
i
n
Ag
r
i
c
u
l
t
u
re
,
v
o
l
.
1
8
8
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
a
g
.
2
0
2
1
.
1
0
6
3
5
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
14
,
No
.
2
,
J
u
n
e
2
0
2
5
:
3
8
4
-
393
392
[
1
5
]
S
.
S
a
i
f
u
l
l
a
h
,
D
.
B
.
P
r
a
se
t
y
o
,
I
n
d
a
h
y
a
n
i
,
R
.
D
r
e
ż
e
w
s
k
i
,
a
n
d
F
.
A
.
D
w
i
y
a
n
t
o
,
“
P
a
l
m
o
i
l
m
a
t
u
r
i
t
y
c
l
a
ss
i
f
i
c
a
t
i
o
n
u
s
i
n
g
k
-
n
e
a
r
e
s
t
n
e
i
g
h
b
o
r
s
b
a
se
d
o
n
R
G
B
a
n
d
l
*
a
*
b
c
o
l
o
r
e
x
t
r
a
c
t
i
o
n
,
”
Pr
o
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
2
2
5
,
p
p
.
3
0
1
1
–
3
0
2
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.
2
0
2
3
.
1
0
.
2
9
4
.
[
1
6
]
H
.
A
l
t
a
h
e
r
i
,
M
.
A
l
s
u
l
a
i
m
a
n
,
a
n
d
G
.
M
u
h
a
mm
a
d
,
“
D
a
t
e
f
r
u
i
t
c
l
a
ssi
f
i
c
a
t
i
o
n
f
o
r
r
o
b
o
t
i
c
h
a
r
v
e
s
t
i
n
g
i
n
a
n
a
t
u
r
a
l
e
n
v
i
r
o
n
me
n
t
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
I
EE
E
A
c
c
e
ss
,
v
o
l
.
7
,
p
p
.
1
1
7
1
1
5
–
1
1
7
1
3
3
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
9
.
2
9
3
6
5
3
6
.
[
1
7
]
M
.
S
.
H
o
s
sai
n
,
M
.
A
l
-
H
a
mm
a
d
i
,
a
n
d
G
.
M
u
h
a
mm
a
d
,
“
A
u
t
o
ma
t
i
c
f
r
u
i
t
c
l
a
ssi
f
i
c
a
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
f
o
r
i
n
d
u
st
r
i
a
l
a
p
p
l
i
c
a
t
i
o
n
s,
”
I
EEE
T
r
a
n
sa
c
t
i
o
n
s
o
n
I
n
d
u
s
t
ri
a
l
I
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
1
5
,
n
o
.
2
,
p
p
.
1
0
2
7
–
1
0
3
4
,
F
e
b
.
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
I
.
2
0
1
8
.
2
8
7
5
1
4
9
.
[
1
8
]
J.
S
t
e
i
n
b
r
e
n
e
r
,
K
.
P
o
s
c
h
,
a
n
d
R
.
L
e
i
t
n
e
r
,
“
H
y
p
e
r
sp
e
c
t
r
a
l
f
r
u
i
t
a
n
d
v
e
g
e
t
a
b
l
e
c
l
a
ssi
f
i
c
a
t
i
o
n
u
s
i
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
,
”
C
o
m
p
u
t
e
rs
a
n
d
E
l
e
c
t
r
o
n
i
c
s
i
n
A
g
r
i
c
u
l
t
u
r
e
,
v
o
l
.
1
6
2
,
p
p
.
3
6
4
–
3
7
2
,
J
u
l
.
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
a
g
.
2
0
1
9
.
0
4
.
0
1
9
.
[
1
9
]
S
.
F
a
n
e
t
a
l
.
,
“
O
n
l
i
n
e
d
e
t
e
c
t
i
o
n
o
f
d
e
f
e
c
t
i
v
e
a
p
p
l
e
s
u
si
n
g
c
o
mp
u
t
e
r
v
i
si
o
n
s
y
st
e
m
c
o
m
b
i
n
e
d
w
i
t
h
d
e
e
p
l
e
a
r
n
i
n
g
me
t
h
o
d
s,
”
J
o
u
rn
a
l
o
f
Fo
o
d
E
n
g
i
n
e
e
ri
n
g
,
v
o
l
.
2
8
6
,
D
e
c
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
f
o
o
d
e
n
g
.
2
0
2
0
.
1
1
0
1
0
2
.
[
2
0
]
M
.
S
e
w
a
k
,
S
.
K
.
S
a
h
a
y
,
a
n
d
H
.
R
a
t
h
o
r
e
,
“
A
n
o
v
e
r
v
i
e
w
o
f
d
e
e
p
l
e
a
r
n
i
n
g
a
r
c
h
i
t
e
c
t
u
r
e
o
f
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s
a
n
d
a
u
t
o
e
n
c
o
d
e
r
s,”
J
o
u
rn
a
l
o
f
C
o
m
p
u
t
a
t
i
o
n
a
l
a
n
d
T
h
e
o
re
t
i
c
a
l
N
a
n
o
s
c
i
e
n
c
e
,
v
o
l
.
1
7
,
n
o
.
1
,
p
p
.
1
8
2
–
1
8
8
,
J
a
n
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
6
6
/
j
c
t
n
.
2
0
2
0
.
8
6
4
8
.
[
2
1
]
S
.
S
.
,
J.
I
.
Z
o
n
g
C
h
e
n
,
a
n
d
S
.
S
h
a
k
y
a
,
“
S
u
r
v
e
y
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
a
r
c
h
i
t
e
c
t
u
r
e
s
w
i
t
h
d
e
e
p
l
e
a
r
n
i
n
g
,
”
J
o
u
r
n
a
l
o
f
S
o
f
t
C
o
m
p
u
t
i
n
g
Pa
r
a
d
i
g
m
,
v
o
l
.
2
,
n
o
.
3
,
p
p
.
1
8
6
–
1
9
4
,
Ju
l
.
2
0
2
0
,
d
o
i
:
1
0
.
3
6
5
4
8
/
j
sc
p
.
2
0
2
0
.
3
.
0
0
7
.
[
2
2
]
N
.
R
a
i
a
n
d
X
.
S
u
n
,
“
W
e
e
d
V
i
si
o
n
:
a
si
n
g
l
e
-
st
a
g
e
d
e
e
p
l
e
a
r
n
i
n
g
a
r
c
h
i
t
e
c
t
u
r
e
t
o
p
e
r
f
o
r
m
w
e
e
d
d
e
t
e
c
t
i
o
n
a
n
d
se
g
men
t
a
t
i
o
n
u
si
n
g
d
r
o
n
e
-
a
c
q
u
i
r
e
d
i
m
a
g
e
s,
”
C
o
m
p
u
t
e
rs
a
n
d
El
e
c
t
ro
n
i
c
s
i
n
A
g
ri
c
u
l
t
u
re
,
v
o
l
.
2
1
9
,
A
p
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
p
a
g
.
2
0
2
4
.
1
0
8
7
9
2
.
[
2
3
]
B
.
B
i
sc
h
l
e
t
a
l
.
,
“
H
y
p
e
r
p
a
r
a
me
t
e
r
o
p
t
i
mi
z
a
t
i
o
n
:
f
o
u
n
d
a
t
i
o
n
s,
a
l
g
o
r
i
t
h
ms
,
b
e
st
p
r
a
c
t
i
c
e
s,
a
n
d
o
p
e
n
c
h
a
l
l
e
n
g
e
s
,
”
WIREs
D
a
t
a
Mi
n
i
n
g
a
n
d
K
n
o
w
l
e
d
g
e
D
i
sc
o
v
e
ry
,
v
o
l
.
1
3
,
n
o
.
2
,
M
a
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
2
/
w
i
d
m.
1
4
8
4
.
[
2
4
]
A
.
R
.
M
a
ma
t
,
F
.
S
.
M
o
h
a
ma
d
,
N
.
a
b
d
u
l
R
a
w
i
,
M
.
K
.
A
w
a
n
g
,
M
.
I
.
A
w
a
n
g
,
a
n
d
M
.
F
.
A
.
K
a
d
i
r
,
“
R
e
g
i
o
n
b
a
se
d
i
m
a
g
e
r
e
t
r
i
e
v
a
l
b
a
s
e
d
o
n
t
e
x
t
u
r
e
f
e
a
t
u
r
e
s
,
”
J
o
u
r
n
a
l
o
f
T
h
e
o
re
t
i
c
a
l
a
n
d
Ap
p
l
i
e
d
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
9
2
,
n
o
.
1
,
2
0
1
6
.
[
2
5
]
K
.
W
a
n
g
c
h
u
k
,
T.
D
o
r
j
i
,
P
.
R
a
m
D
h
u
n
g
y
e
l
,
a
n
d
P
.
G
a
l
e
y
,
“
F
r
u
i
t
s
a
n
d
v
e
g
e
t
a
b
l
e
s
r
e
c
o
g
n
i
t
i
o
n
s
y
s
t
e
m
i
n
D
z
o
n
g
k
h
a
u
si
n
g
v
i
s
u
a
l
g
e
o
me
t
r
y
g
r
o
u
p
n
e
t
w
o
r
k
,
”
Z
o
r
i
g
Me
l
o
n
g
-
A
T
e
c
h
n
i
c
a
l
J
o
u
r
n
a
l
o
f
S
c
i
e
n
c
e
,
En
g
i
n
e
e
ri
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
8
,
n
o
.
1
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
1
7
1
0
2
/
v
8
0
0
4
.
[
2
6
]
M
.
M
u
k
h
i
d
d
i
n
o
v
,
A
.
M
u
m
i
n
o
v
,
a
n
d
J
.
C
h
o
,
“
I
mp
r
o
v
e
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
p
p
r
o
a
c
h
f
o
r
f
r
u
i
t
s
a
n
d
v
e
g
e
t
a
b
l
e
s
f
r
e
s
h
n
e
s
s
b
a
se
d
o
n
d
e
e
p
l
e
a
r
n
i
n
g
,
”
S
e
n
s
o
rs
,
v
o
l
.
2
2
,
n
o
.
2
1
,
O
c
t
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
2
2
1
8
1
9
2
.
[
2
7
]
C.
-
H
.
C
h
a
n
g
,
L.
R
a
m
p
a
s
e
k
,
a
n
d
A
.
G
o
l
d
e
n
b
e
r
g
,
“
D
r
o
p
o
u
t
f
e
a
t
u
r
e
r
a
n
k
i
n
g
f
o
r
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
s,
”
a
r
Xi
v
,
D
e
c
.
2
0
1
7
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
a
r
x
i
v
.
o
r
g
/
a
b
s
/
1
7
1
2
.
0
8
6
4
5
[
2
8
]
D
.
P
.
K
i
n
g
ma
a
n
d
J
.
B
a
,
“
A
d
a
m:
a
me
t
h
o
d
f
o
r
s
t
o
c
h
a
s
t
i
c
o
p
t
i
mi
z
a
t
i
o
n
,
”
a
rX
i
v
,
D
e
c
.
2
0
1
4
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
a
r
x
i
v
.
o
r
g
/
a
b
s/
1
4
1
2
.
6
9
8
0
[
2
9
]
A
.
F
.
A
g
a
r
a
p
,
“
D
e
e
p
l
e
a
r
n
i
n
g
u
s
i
n
g
r
e
c
t
i
f
i
e
d
l
i
n
e
a
r
u
n
i
t
s
(
R
e
LU
)
,
”
a
rX
i
v
,
M
a
r
.
2
0
1
8
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
a
r
x
i
v
.
o
r
g
/
a
b
s/
1
8
0
3
.
0
8
3
7
5
[
3
0
]
R
.
A
r
o
r
a
,
A
.
B
a
su
,
P
.
M
i
a
n
j
y
,
a
n
d
A
.
M
u
k
h
e
r
j
e
e
,
“
U
n
d
e
r
s
t
a
n
d
i
n
g
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
s
w
i
t
h
r
e
c
t
i
f
i
e
d
l
i
n
e
a
r
u
n
i
t
s,
”
a
rXi
v
,
N
o
v
.
2
0
1
6
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
a
r
x
i
v
.
o
r
g
/
a
b
s/
1
6
1
1
.
0
1
4
9
1
[
3
1
]
S
.
R
.
D
u
b
e
y
,
S
.
K
.
S
i
n
g
h
,
a
n
d
B
.
B
.
C
h
a
u
d
h
u
r
i
,
“
A
c
t
i
v
a
t
i
o
n
f
u
n
c
t
i
o
n
s i
n
d
e
e
p
l
e
a
r
n
i
n
g
:
a
c
o
m
p
r
e
h
e
n
si
v
e
su
r
v
e
y
a
n
d
b
e
n
c
h
m
a
r
k
,
”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
5
0
3
,
p
p
.
9
2
–
1
0
8
,
S
e
p
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m
.
2
0
2
2
.
0
6
.
1
1
1
.
[
3
2
]
C
.
G
a
r
b
i
n
,
X
.
Z
h
u
,
a
n
d
O
.
M
a
r
q
u
e
s,
“
D
r
o
p
o
u
t
v
s
.
b
a
t
c
h
n
o
r
ma
l
i
z
a
t
i
o
n
:
a
n
e
m
p
i
r
i
c
a
l
s
t
u
d
y
o
f
t
h
e
i
r
i
m
p
a
c
t
t
o
d
e
e
p
l
e
a
r
n
i
n
g
,
”
Mu
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
7
9
,
n
o
.
1
9
–
2
0
,
p
p
.
1
2
7
7
7
–
1
2
8
1
5
,
M
a
y
2
0
2
0
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
0
4
2
-
0
1
9
-
0
8
4
5
3
-
9.
[
3
3
]
N
.
S
r
i
v
a
s
t
a
v
a
,
G
.
H
i
n
t
o
n
,
A
.
K
r
i
z
h
e
v
s
k
y
,
I
.
S
u
t
s
k
e
v
e
r
,
a
n
d
R
.
S
a
l
a
k
h
u
t
d
i
n
o
v
,
“
D
r
o
p
o
u
t
:
a
s
i
mp
l
e
w
a
y
t
o
p
r
e
v
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
r
o
m
o
v
e
r
f
i
t
t
i
n
g
,
”
J
o
u
rn
a
l
o
f
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
R
e
se
a
rc
h
,
v
o
l
.
1
5
,
p
p
.
1
9
2
9
–
1
9
5
8
,
2
0
1
4
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
j
m
l
r
.
o
r
g
/
p
a
p
e
r
s/
v
o
l
u
me
1
5
/
sr
i
v
a
st
a
v
a
1
4
a
/
sr
i
v
a
s
t
a
v
a
1
4
a
.
p
d
f
[
3
4
]
S
.
K
.
D
e
v
a
l
l
a
e
t
a
l
.
,
“
A
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
t
o
d
i
g
i
t
a
l
l
y
s
t
a
i
n
o
p
t
i
c
a
l
c
o
h
e
r
e
n
c
e
t
o
m
o
g
r
a
p
h
y
i
ma
g
e
s
o
f
t
h
e
o
p
t
i
c
n
e
r
v
e
h
e
a
d
,
”
I
n
v
e
s
t
i
g
a
t
i
v
e
O
p
t
h
a
l
m
o
l
o
g
y
& V
i
s
u
a
l
S
c
i
e
n
c
e
,
v
o
l
.
5
9
,
n
o
.
1
,
Ja
n
.
2
0
1
8
,
d
o
i
:
1
0
.
1
1
6
7
/
i
o
v
s.
1
7
-
2
2
6
1
7
.
[
3
5
]
M
.
G
.
A
l
i
y
u
,
M
.
F
.
A
.
K
a
d
i
r
,
A
.
R
.
M
a
m
a
t
,
a
n
d
M
.
M
o
h
a
m
a
d
,
“
N
o
i
s
e
r
e
m
o
v
a
l
u
si
n
g
st
a
t
i
s
t
i
c
a
l
o
p
e
r
a
t
o
r
s
f
o
r
e
f
f
i
c
i
e
n
t
l
e
a
f
i
d
e
n
t
i
f
i
c
a
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
Ai
d
e
d
En
g
i
n
e
e
ri
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
0
,
n
o
.
4
,
2
0
1
8
,
d
o
i
:
1
0
.
1
5
0
4
/
I
JC
A
ET
.
2
0
1
8
.
0
9
2
8
3
4
.
[
3
6
]
Ò
.
Lo
r
e
n
t
e
,
I
.
R
i
e
r
a
,
a
n
d
A
.
R
a
n
a
,
“
I
mag
e
c
l
a
ss
i
f
i
c
a
t
i
o
n
w
i
t
h
c
l
a
ss
i
c
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
a
r
Xi
v
,
M
a
y
2
0
2
1
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
a
r
x
i
v
.
o
r
g
/
a
b
s
/
2
1
0
5
.
0
4
8
9
5
[
3
7
]
M
.
R
i
z
o
n
,
N
.
A
.
N
a
j
i
h
a
h
Y
u
sr
i
,
M
.
F
.
A
b
d
u
l
K
a
d
i
r
,
A
.
R
.
b
i
n
M
a
m
a
t
,
A
.
Z.
A
b
d
A
z
i
z
,
a
n
d
K
.
N
a
n
a
a
,
“
D
e
t
e
r
m
i
n
a
t
i
o
n
o
f
ma
n
g
o
f
r
u
i
t
f
r
o
m
b
i
n
a
r
y
i
ma
g
e
u
si
n
g
r
a
n
d
o
m
i
z
e
d
h
o
u
g
h
t
r
a
n
sf
o
r
m,”
i
n
E
i
g
h
t
h
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Ma
c
h
i
n
e
Vi
s
i
o
n
(
I
C
MV
2
0
1
5
)
,
A
.
V
e
r
i
k
a
s
,
P
.
R
a
d
e
v
a
,
a
n
d
D
.
N
i
k
o
l
a
e
v
,
E
d
s.
,
D
e
c
.
2
0
1
5
,
d
o
i
:
1
0
.
1
1
1
7
/
1
2
.
2
2
2
8
5
1
1
.
[
3
8
]
E.
C
.
T
o
o
,
L.
Y
u
j
i
a
n
,
S
.
N
j
u
k
i
,
a
n
d
L.
Y
i
n
g
c
h
u
n
,
“
A
c
o
m
p
a
r
a
t
i
v
e
s
t
u
d
y
o
f
f
i
n
e
-
t
u
n
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
mo
d
e
l
s
f
o
r
p
l
a
n
t
d
i
s
e
a
s
e
i
d
e
n
t
i
f
i
c
a
t
i
o
n
,
”
C
o
m
p
u
t
e
rs
a
n
d
El
e
c
t
r
o
n
i
c
s
i
n
A
g
r
i
c
u
l
t
u
r
e
,
v
o
l
.
1
6
1
,
p
p
.
2
7
2
–
2
7
9
,
J
u
n
.
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
p
a
g
.
2
0
1
8
.
0
3
.
0
3
2
.
[
3
9
]
S
.
C
h
e
t
l
u
r
e
t
a
l
.
,
“
C
u
D
N
N
:
e
f
f
i
c
i
e
n
t
p
r
i
mi
t
i
v
e
s
f
o
r
d
e
e
p
l
e
a
r
n
i
n
g
,
”
a
rX
i
v
,
O
c
t
.
2
0
1
4
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
a
r
x
i
v
.
o
r
g
/
a
b
s/
1
4
1
0
.
0
7
5
9
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Abd
Ra
si
d
Ma
m
a
t
re
c
e
iv
e
d
P
h
.
D
in
Co
m
p
u
ter
S
c
ien
c
e
at
Un
iv
e
rsiti
S
u
lt
a
n
Zain
a
l
Ab
i
d
in
in
2
0
2
2
,
wh
e
re
h
e
c
u
rre
n
tl
y
se
rv
e
s
a
s
a
S
e
n
i
o
r
Lec
tu
re
r.
His
m
a
in
re
se
a
rc
h
a
re
a
s
in
c
lu
d
e
ima
g
e
p
r
o
c
e
ss
in
g
,
c
o
n
ten
t
-
b
a
se
d
ima
g
e
re
tri
e
v
a
l,
d
e
c
isio
n
su
p
p
o
rt
sy
ste
m
s,
a
n
d
d
e
e
p
lea
rn
in
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
rm
@u
n
isz
a
.
e
d
u
.
m
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
C
o
mp
a
r
a
tive
s
tu
d
y
o
n
fin
e
-
t
u
n
in
g
d
ee
p
lea
r
n
in
g
mo
d
els fo
r
fr
u
it a
n
d
ve
g
eta
b
le
…
(
A
b
d
R
a
s
id
Ma
ma
t)
393
Mo
h
a
m
a
d
Afe
n
d
e
e
M
o
h
a
m
e
d
re
c
e
iv
e
d
h
is
P
h
.
D
in
M
a
th
e
m
a
ti
c
a
l
Cry
p
t
o
g
ra
p
h
y
at
Un
iv
e
rsiti
P
u
tra M
a
lay
sia
in
2
0
1
1
.
Up
o
n
c
o
m
p
letio
n
,
h
e
se
rv
e
d
t
h
e
u
n
i
v
e
rsity
fo
r
th
re
e
y
e
a
rs
a
s a
se
n
io
r
lec
tu
re
r.
In
2
0
1
4
,
h
e
m
o
v
e
d
t
o
Un
i
v
e
rsiti
S
u
lt
a
n
Zain
a
l
Ab
id
i
n
a
n
d
late
r
a
ss
u
m
e
d
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
p
o
si
ti
o
n
.
Hi
s
c
u
rre
n
t
re
se
a
rc
h
in
tere
sts
i
n
c
lu
d
e
b
o
th
th
e
o
re
ti
c
a
l
a
n
d
a
p
p
li
c
a
ti
o
n
issu
e
s
in
th
e
d
o
m
a
in
o
f
d
a
ta
se
c
u
rit
y
a
n
d
m
o
b
il
e
a
n
d
wire
les
s
n
e
two
r
k
in
g
.
Dr
M
o
h
a
m
e
d
h
a
s
a
u
th
o
re
d
m
o
re
th
a
n
1
0
0
a
rti
c
les
th
a
t
h
a
v
e
a
p
p
e
a
r
e
d
in
v
a
rio
u
s
jo
u
rn
a
ls,
b
o
o
k
c
h
a
p
ters
,
a
n
d
c
o
n
fe
re
n
c
e
p
ro
c
e
e
d
in
g
s.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
m
a
fe
n
d
e
e
@u
n
isz
a
.
e
d
u
.
m
y
.
Mo
h
d
Fa
d
z
il
Abd
u
l
K
a
d
ir
re
c
e
iv
e
d
P
h
.
D
.
in
En
g
in
e
e
rin
g
(S
y
s
tem
En
g
in
e
e
ri
n
g
)
at
th
e
M
ie
Un
iv
e
rsity
,
M
ie,
Ja
p
a
n
,
in
2
0
1
2
.
S
i
n
c
e
2
0
0
6
,
h
e
h
a
s
b
e
e
n
with
th
e
F
a
c
u
l
ty
o
f
In
fo
rm
a
ti
c
s
a
n
d
C
o
m
p
u
ti
n
g
,
Un
i
v
e
rsiti
S
u
lt
a
n
Zai
n
a
l
Ab
i
d
in
,
wh
e
re
h
e
is
c
u
rre
n
tl
y
a
S
e
n
io
r
Lec
tu
re
r.
His
m
a
in
a
re
a
s
o
f
re
se
a
rc
h
in
tere
st
a
re
d
ig
it
a
l
ima
g
e
p
ro
c
e
ss
in
g
,
p
a
t
ter
n
re
c
o
g
n
it
io
n
,
in
fo
rm
a
ti
o
n
se
c
u
rit
y
,
a
n
d
c
ry
p
t
o
g
ra
p
h
y
.
He
is
a
lso
a
m
e
m
b
e
r
o
f
th
e
M
a
lay
sia
Bo
a
rd
o
f
Tec
h
n
o
l
o
g
ists.
H
e
c
a
n
b
e
c
o
n
tac
t
e
d
a
t
e
m
a
il
:
fa
d
z
il
@u
n
isz
a
.
e
d
u
.
m
y
.
No
r
k
h
a
ira
n
i
Abd
u
l
Ra
wi
is
a
se
n
io
r
lec
tu
re
r
in
th
e
De
p
a
rtme
n
t
o
f
M
u
lt
ime
d
ia
S
tu
d
ies
,
F
a
c
u
lt
y
o
f
I
n
fo
rm
a
ti
c
s
a
n
d
Co
m
p
u
ti
n
g
,
U
n
iv
e
rsiti
S
u
lt
a
n
Zain
a
l
Ab
id
i
n
.
S
h
e
re
c
e
iv
e
d
h
e
r
m
a
ste
r’s
d
e
g
re
e
in
i
n
fo
rm
a
ti
o
n
sc
ien
c
e
in
2
0
0
3
.
He
r
re
se
a
rc
h
in
tere
sts
a
re
m
u
lt
ime
d
ia
a
n
d
ima
g
e
p
ro
c
e
ss
in
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
k
h
a
iran
i@
u
n
isz
a
.
e
d
u
.
m
y
.
Az
im
Za
l
ih
a
Abd
Az
iz
is
a
se
n
io
r
lec
tu
re
r
at
t
h
e
F
a
c
u
lt
y
o
f
In
f
o
rm
a
ti
c
s
a
n
d
Co
m
p
u
ti
n
g
,
Un
i
v
e
rsiti
S
u
lt
a
n
Zain
a
l
Ab
i
d
in
(U
n
iS
ZA),
Tere
n
g
g
a
n
u
,
M
a
lay
sia
.
S
h
e
re
c
e
iv
e
d
h
e
r
P
h
D
(C
o
m
p
u
ter
S
c
ien
c
e
)
fro
m
th
e
Un
i
v
e
rsity
o
f
Re
a
d
i
n
g
,
Un
it
e
d
Kin
g
d
o
m
in
2
0
1
7
.
S
h
e
d
id
h
e
r
first
d
e
g
re
e
in
Co
m
p
u
t
e
r
S
c
ien
c
e
in
2
0
0
0
a
t
th
e
F
a
c
u
lt
y
o
f
Co
m
p
u
ter
S
c
ie
n
c
e
,
Un
iv
e
rsiti
Te
k
n
o
lo
g
i
M
a
lay
sia
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
z
imz
a
li
h
a
@u
n
isz
a
.
e
d
u
.
m
y
.
Wa
n
S
u
r
y
a
n
i
Wa
n
Aw
a
n
g
is
a
se
n
io
r
lec
tu
re
r
i
n
Co
m
p
u
ter
S
c
i
e
n
c
e
a
t
Un
i
v
e
rsiti
S
u
lt
a
n
Zai
n
a
l
Ab
id
i
n
in
Ku
a
la
T
e
re
n
g
g
a
n
u
,
M
a
lay
sia
,
wh
e
re
sh
e
h
a
s
wo
r
k
e
d
si
n
c
e
2
0
0
7
.
S
h
e
e
a
rn
e
d
h
e
r
u
n
d
e
rg
ra
d
u
a
te
d
e
g
re
e
in
c
o
m
p
u
ter
stu
d
ies
fro
m
S
h
e
ffie
ld
Ha
ll
a
m
Un
iv
e
rsity
i
n
th
e
UK
,
a
m
a
ste
r’s
d
e
g
re
e
fro
m
wh
a
t
is
n
o
w
Un
iv
e
rsiti
M
a
lay
sia
T
e
re
n
g
g
a
n
u
,
a
n
d
a
P
h
.
D.
in
Co
m
p
u
ter
S
c
ien
c
e
fro
m
Ca
rd
iff
Un
iv
e
rsity
.
He
r
e
x
p
e
rti
se
in
d
istri
b
u
ted
s
y
ste
m
s
a
n
d
d
a
tab
a
se
s
a
ll
o
ws
h
e
r
to
o
ffe
r
v
a
l
u
a
b
le
in
si
g
h
ts
i
n
b
o
th
a
c
a
d
e
m
ic
a
n
d
p
ro
fe
ss
io
n
a
l
se
tt
in
g
s.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
su
r
y
a
n
i@u
n
is
z
a
.
e
d
.
m
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.