I
AE
S
I
nte
rna
t
io
na
l J
o
urna
l o
f
Ro
bo
t
ics a
nd
Aut
o
m
a
t
io
n (
I
J
RA)
Vo
l.
1
0
,
No
.
2
,
J
u
n
e
2
0
2
1
,
p
p
.
1
2
3
~1
3
2
I
SS
N:
2722
-
2586
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
r
a
.
v
1
0
i
2
.
p
p
1
2
3
-
132
123
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
a
.
ia
esco
r
e.
co
m
M
iss
ing
data
hand
ling
f
o
r
m
a
chin
e
learning
m
o
dels
K
a
ri
m
H
.
E
ria
n,
P
edro
H
.
Reg
a
la
do
,
J
a
m
es M
.
Co
nra
d
ECE
De
p
a
rtm
e
n
t,
Un
iv
e
rsit
y
o
f
No
rth
Ca
ro
li
n
a
a
t
C
h
a
rlo
t
te,
Ch
a
r
lo
tt
e
NC,
U
n
it
e
d
S
tate
s
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
1
9
,
2
0
2
0
R
ev
i
s
ed
Dec
1
,
2
0
2
0
A
cc
ep
ted
Feb
1
2
,
2
0
2
1
T
h
is
p
a
p
e
r
d
isc
u
ss
e
s
a
n
o
v
e
l
a
lg
o
rit
h
m
f
o
r
so
lv
in
g
a
m
issin
g
d
a
ta
p
ro
b
lem
in
th
e
m
a
c
h
in
e
lea
rn
i
n
g
p
re
-
p
r
o
c
e
ss
in
g
sta
g
e
.
A
m
o
d
e
l
b
u
il
t
to
h
e
lp
len
d
e
rs
e
v
a
lu
a
te
h
o
m
e
lo
a
n
s
b
a
se
d
o
n
n
u
m
e
ro
u
s
f
a
c
to
rs
b
y
lea
rn
in
g
f
ro
m
a
v
a
il
a
b
le
u
se
r
d
a
ta,
is
a
d
o
p
ted
in
t
h
is
p
a
p
e
r
a
s
a
n
e
x
a
m
p
le.
If
o
n
e
o
f
th
e
f
a
c
to
rs
is
m
is
sin
g
f
o
r
a
p
e
rso
n
in
t
h
e
d
a
ta
se
t,
th
e
c
u
rre
n
tl
y
u
se
d
m
e
th
o
d
s
d
e
lete
th
e
w
h
o
l
e
e
n
try
th
e
re
f
o
re
re
d
u
c
in
g
th
e
siz
e
o
f
th
e
d
a
tas
e
t
a
n
d
a
f
f
e
c
ti
n
g
th
e
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
l
a
c
c
u
ra
c
y
.
T
h
e
n
o
v
e
l
a
lg
o
rit
h
m
a
ims
to
a
v
o
id
l
o
sin
g
e
n
tri
e
s
f
o
r
m
issin
g
f
a
c
to
rs
b
y
b
re
a
k
in
g
th
e
d
a
tas
e
t
in
to
m
u
lt
ip
l
e
su
b
se
ts,
b
u
il
d
in
g
a
d
if
f
e
r
e
n
t
m
a
c
h
in
e
lea
r
n
in
g
m
o
d
e
l
f
o
r
e
a
c
h
su
b
se
t,
th
e
n
c
o
m
b
in
in
g
th
e
m
o
d
e
ls
i
n
to
o
n
e
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
l.
In
t
h
is
m
a
n
n
e
r,
t
h
e
m
o
d
e
l
m
a
k
e
s
u
se
o
f
a
ll
a
v
a
il
a
b
le
d
a
ta
a
n
d
o
n
ly
n
e
g
le
c
ts
th
e
m
issin
g
v
a
lu
e
s.
Ov
e
ra
ll
,
th
e
n
e
w
a
lg
o
rit
h
m
im
p
ro
v
e
d
th
e
p
re
d
icti
o
n
a
c
c
u
ra
c
y
b
y
5
%
fro
m
9
3
%
acc
u
ra
c
y
to
9
8
%
in
th
e
h
o
m
e
lo
a
n
e
x
a
m
p
le.
K
ey
w
o
r
d
s
:
C
las
s
i
f
icatio
n
Data
p
r
e
-
p
r
o
ce
s
s
in
g
Featu
r
e
s
elec
t
io
n
L
o
g
i
s
tic
r
eg
r
e
s
s
io
n
Ma
ch
i
n
e
lear
n
i
n
g
T
ar
g
et
W
eig
h
t
s
Fi
f
t
h
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
:
Kar
i
m
H.
E
r
ian
E
C
E
Dep
ar
t
m
en
t
Un
i
v
er
s
it
y
o
f
No
r
th
C
ar
o
lin
a
a
t Ch
ar
lo
tte
9
2
0
1
Un
iv
er
s
it
y
C
it
y
B
lv
d
,
C
h
ar
lo
tte,
NC
2
8
2
2
3
,
U
n
ited
States
E
m
ail:
k
er
ian
@
u
n
cc
.
ed
u
1.
I
NT
RO
D
UCT
I
O
N
I
n
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
i
n
g
m
o
d
els,
t
h
e
ac
c
u
r
ac
y
o
f
t
h
e
m
o
d
el
d
ep
en
d
s
o
n
th
e
q
u
alit
y
o
f
th
e
av
ailab
le
d
ataset
u
s
ed
f
o
r
tr
ain
in
g
.
I
n
m
o
s
t
d
atasets
,
th
er
e
ar
e
m
i
s
s
i
n
g
ite
m
s
.
T
h
e
w
a
y
o
f
h
a
n
d
lin
g
th
e
m
is
s
in
g
ite
m
s
is
cr
itical
to
tr
ai
n
t
h
e
m
o
d
el
to
w
ar
d
g
et
ti
n
g
t
h
e
ex
p
ec
ted
r
esu
lts
.
T
h
e
m
o
s
t
co
m
m
o
n
w
a
y
s
to
h
an
d
l
e
m
is
s
i
n
g
ite
m
s
ar
e
b
y
r
e
m
o
v
a
l
o
f
th
e
w
h
o
le
en
tr
y
,
b
y
r
ep
lacin
g
t
h
e
m
i
s
s
i
n
g
ite
m
s
b
y
it
s
d
ef
au
lt
v
al
u
e,
b
y
ch
an
g
i
n
g
t
h
e
m
is
s
i
n
g
ite
m
s
to
a
ze
r
o
,
o
r
b
y
w
r
i
tin
g
t
h
e
av
er
a
g
e
v
al
u
e
i
n
s
tead
.
I
n
t
h
is
p
ap
er
,
th
e
tea
m
u
s
ed
an
e
x
a
m
p
le
o
f
h
o
m
e
lo
an
s
d
ataset
w
h
ic
h
i
n
cl
u
d
es
2
2
0
f
ie
ld
s
an
d
b
y
r
e
m
o
v
i
n
g
th
e
m
is
s
in
g
ite
m
s
,
t
h
e
d
ataset
s
h
r
u
n
k
f
r
o
m
3
0
7
,
5
1
1
en
tr
ies
to
8
,
6
0
3
en
tr
ies
a
n
d
a
9
7
.
2
%
r
ed
u
ctio
n
in
d
ata
s
ize
af
f
ec
ted
t
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
e
s
ti
m
atio
n
.
T
h
e
o
th
er
s
o
lu
tio
n
s
m
a
n
ip
u
late
t
h
e
d
ata
b
y
in
tr
o
d
u
ci
n
g
an
o
t
h
er
v
alu
e
t
h
a
t
is
n
o
t
th
e
r
ea
l
v
alu
e
f
o
r
th
is
en
tr
y
,
lead
in
g
also
to
lo
w
er
ac
cu
r
ac
y
[
1
]
.
T
h
is
h
u
g
e
r
ed
u
ctio
n
u
r
g
ed
th
e
team
to
f
i
n
d
an
o
th
er
w
a
y
to
h
an
d
le
th
e
m
i
s
s
i
n
g
en
tr
ies.
T
h
e
tea
m
d
ec
id
ed
to
co
m
e
u
p
w
it
h
a
w
a
y
to
p
ay
atte
n
tio
n
to
ev
er
y
u
n
iq
u
e
ce
ll
th
at
is
b
lan
k
,
in
s
tead
o
f
j
u
s
t
d
eletin
g
a
w
h
o
le
r
o
w
i
f
o
n
e
ce
ll
is
m
is
s
i
n
g
d
ata.
E
s
p
ec
ially
w
h
e
n
th
e
tea
m
w
a
s
a
w
ar
e
t
h
at
th
er
e
w
er
e
2
2
0
u
n
iq
u
e
f
ea
tu
r
es
in
t
h
e
d
ataset,
an
d
h
o
w
h
ar
d
it
w
as to
g
a
th
er
m
o
r
e
d
ata.
T
h
e
n
o
v
el
alg
o
r
it
h
m
p
r
o
p
o
s
ed
in
th
i
s
p
ap
er
ai
m
s
to
o
n
l
y
r
e
m
o
v
e
t
h
e
m
is
s
in
g
ce
lls
.
T
h
is
c
an
b
e
d
o
n
e
b
y
b
r
ea
k
i
n
g
t
h
e
d
ataset
in
to
m
u
ltip
le
s
u
b
s
ets,
ea
ch
co
n
s
i
s
ti
n
g
o
f
o
n
l
y
o
n
e
f
ea
t
u
r
e.
E
ac
h
s
u
b
s
et
is
u
s
ed
to
tr
ain
a
s
i
m
p
le
m
ac
h
i
n
e
lear
n
i
n
g
m
o
d
el.
T
h
en
u
s
i
n
g
a
w
ei
g
h
ti
n
g
m
ec
h
a
n
is
m
,
all
m
o
d
el
s
ar
e
co
m
b
i
n
ed
in
to
o
n
e
m
o
d
el
t
h
at
u
s
es a
l
l th
e
a
v
ailab
le
f
ea
t
u
r
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2586
I
A
E
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
10
,
No
.
2
,
J
u
n
e
2
0
2
1
:
1
2
3
–
132
124
2.
T
H
E
P
RO
P
O
SE
D
AL
G
O
RI
T
H
M
AND
A
SURVE
Y
O
F
E
XI
ST
I
N
G
M
E
T
H
O
DS
W
h
en
r
esear
c
h
i
n
g
w
a
y
s
to
h
a
n
d
le
m
is
s
in
g
d
ata,
t
h
e
tea
m
f
o
u
n
d
t
h
at
o
t
h
er
s
h
av
e
tr
ied
a
v
ar
iet
y
o
f
d
if
f
er
e
n
t
m
eth
o
d
s
,
b
u
t
o
n
l
y
t
h
e
to
p
f
iv
e
m
et
h
o
d
s
f
o
r
h
a
n
d
li
n
g
m
i
s
s
i
n
g
d
ata
w
il
l
b
e
d
is
cu
s
s
ed
[
2
-
7
]
.
I
n
ea
ch
m
et
h
o
d
,
ad
v
an
tag
e
s
an
d
d
is
ad
v
an
ta
g
e
s
ar
e
s
tated
.
T
h
e
f
iv
e
m
et
h
o
d
s
f
o
u
n
d
in
r
esear
c
h
ar
e:
‑
d
eletin
g
r
o
w
s
,
‑
r
ep
lacin
g
w
it
h
m
ea
n
o
r
m
ed
ia
n
o
r
m
o
d
e
,
‑
ass
i
g
n
in
g
a
u
n
iq
u
e
ca
te
g
o
r
y
,
‑
p
r
ed
ictin
g
t
h
e
m
i
s
s
i
n
g
v
al
u
es
,
an
d
‑
u
s
i
n
g
alg
o
r
it
h
m
s
t
h
at
s
u
p
p
o
r
t
m
is
s
i
n
g
v
al
u
es
.
2
.
1
.
Dele
t
ing
r
o
w
s
I
n
th
e
f
ir
s
t
m
et
h
o
d
,
th
e
ad
v
a
n
tag
e
s
o
f
d
eleti
n
g
r
o
w
s
ca
n
r
esu
lt
i
n
a
r
o
b
u
s
t
a
n
d
ac
cu
r
a
te
m
o
d
el.
Dis
ad
v
a
n
ta
g
es
o
f
d
eleti
n
g
r
o
w
s
i
s
th
e
lo
s
s
o
f
in
f
o
r
m
at
io
n
an
d
d
ata.
T
h
e
b
ig
g
est
ch
al
len
g
e
i
s
th
at
i
f
th
e
m
is
s
i
n
g
v
al
u
es a
r
e
a
b
ig
p
er
ce
n
tag
e
o
f
th
e
d
ata,
t
h
en
t
h
e
d
ataset
is
b
o
u
n
d
to
b
e
s
k
e
w
ed
.
2
.
2
.
Repla
cing
w
it
h
m
ea
n o
r
m
e
dia
n o
r
m
o
de
I
n
th
e
s
ec
o
n
d
m
e
th
o
d
,
ad
v
an
t
ag
es
o
f
r
ep
lacin
g
w
it
h
th
e
m
ea
n
,
m
ed
ian
,
o
r
m
o
d
e
is
th
at
it
ca
n
ad
d
v
ar
ian
ce
to
th
e
d
ataset.
T
h
is
c
an
b
e
g
o
o
d
w
h
en
t
h
e
d
ataset
i
s
s
m
all,
a
n
d
it
also
p
r
ev
e
n
t
s
d
ata
lo
s
s
o
r
r
e
m
o
v
al.
Dis
ad
v
a
n
ta
g
es
o
f
r
ep
lacin
g
w
i
th
th
e
m
ea
n
,
m
ed
ian
,
o
r
m
o
d
e
is
th
at
i
m
p
u
ti
n
g
th
e
v
ar
iatio
n
s
ad
d
v
ar
ian
ce
an
d
b
ias.
A
l
s
o
,
th
er
e
ar
e
m
o
r
e
co
m
p
lex
i
m
p
u
tatio
n
m
et
h
o
d
s
th
at
ar
e
b
etter
,
y
et
m
o
r
e
co
m
p
le
x
.
2
.
3
.
Ass
ig
nin
g
a
un
iqu
e
ca
t
eg
o
ry
I
n
t
h
e
t
h
ir
d
m
et
h
o
d
,
ad
v
a
n
tag
es
o
f
a
s
s
i
g
n
i
n
g
a
u
n
iq
u
e
ca
te
g
o
r
y
ar
e
t
h
at
t
h
er
e
ar
e
le
s
s
p
o
s
s
ib
ilit
ie
s
w
it
h
ad
d
in
g
o
n
e
e
x
tr
a
ca
te
g
o
r
y
,
th
is
r
es
u
lt
s
i
n
lo
w
v
ar
ia
n
ce
.
Dis
ad
v
a
n
ta
g
es
o
f
ass
ig
n
i
n
g
a
u
n
iq
u
e
ca
te
g
o
r
y
i
s
th
at
it
ad
d
s
le
s
s
v
ar
ian
ce
a
n
d
also
th
at
b
y
ad
d
i
n
g
a
n
o
th
er
f
ea
t
u
r
e
to
th
e
m
o
d
el
w
h
ile
e
n
co
d
in
g
m
a
y
h
av
e
n
eg
at
iv
e
r
es
u
lt
s
in
t
h
e
o
r
ig
i
n
al
d
ata
an
d
in
r
etu
r
n
g
i
v
e
a
p
o
o
r
ac
cu
r
ac
y
a
n
d
p
er
f
o
r
m
a
n
ce
.
2
.
4
.
P
re
dict
ing
t
he
m
i
s
s
i
ng
v
a
lues
I
n
th
e
f
o
u
r
t
h
m
et
h
o
d
,
ad
v
an
t
ag
es
o
f
p
r
ed
icti
n
g
t
h
e
m
i
s
s
i
n
g
v
al
u
e
s
ar
e
th
at
i
m
p
u
ti
n
g
t
h
e
m
is
s
i
n
g
v
ar
iab
le
ca
n
b
e
an
i
m
p
r
o
v
e
m
en
t
as
lo
n
g
a
s
t
h
e
b
ias
f
r
o
m
t
h
e
s
a
m
e
is
s
m
aller
t
h
an
t
h
e
o
m
itted
v
ar
iab
le
b
ias.
Dis
ad
v
a
n
ta
g
es
o
f
p
r
ed
ictin
g
m
is
s
i
n
g
v
al
u
es
is
t
h
at
it
is
o
n
l
y
tr
u
l
y
co
n
s
id
er
ed
a
s
u
b
s
ti
tu
te
f
o
r
th
e
tr
u
e
m
i
s
s
i
n
g
v
alu
e
s
.
A
ls
o
,
t
h
e
b
ias ca
n
also
in
cr
ea
s
e
w
h
en
a
n
in
co
m
p
lete
c
o
n
d
itio
n
i
n
g
s
et
is
u
s
ed
f
o
r
ca
te
g
o
r
ical
v
al
u
es.
2
.
5
.
Usi
ng
a
lg
o
rit
h
m
s
t
ha
t
s
up
po
rt
m
is
s
ing
v
a
lues
I
n
th
e
f
i
f
t
h
m
et
h
o
d
,
ad
v
an
tag
es
o
f
u
s
i
n
g
alg
o
r
it
h
m
s
th
a
t
s
u
p
p
o
r
t
m
i
s
s
i
n
g
v
al
u
es
is
th
at
t
h
e
y
d
o
n
o
t
r
eq
u
ir
e
cr
ea
tio
n
o
f
a
p
r
e
d
ictiv
e
m
o
d
el
f
o
r
ea
ch
ce
ll
w
it
h
m
i
s
s
i
n
g
d
ata
in
t
h
e
d
ataset.
Als
o
,
it
n
eg
ates
t
h
e
lo
s
s
o
f
d
ata
b
y
ad
d
in
g
a
u
n
iq
u
e
ca
t
eg
o
r
y
.
Dis
ad
v
an
ta
g
es
o
f
u
s
in
g
alg
o
r
ith
m
s
t
h
at
s
u
p
p
o
r
t
m
is
s
i
n
g
v
al
u
es
i
s
th
at
i
t
ad
d
s
less
v
ar
ian
ce
a
n
d
ad
d
s
an
o
th
er
f
ea
tu
r
e
to
th
e
m
o
d
el
w
h
ile
e
n
co
d
in
g
,
a
n
d
th
er
ef
o
r
e
m
a
y
c
h
alle
n
g
e
t
h
e
p
er
f
o
r
m
a
n
ce
a
n
d
ac
cu
r
ac
y
.
2
.
6
.
Su
rv
ey
o
f
re
la
t
ed
wo
rk
I
n
th
e
‘
B
ac
k
g
r
o
u
n
d
’
s
ec
tio
n
,
th
er
e
w
er
e
f
i
v
e
d
if
f
er
en
t
m
et
h
o
d
s
d
is
cu
s
s
ed
th
at
w
er
e
r
esear
ch
ed
f
o
r
d
if
f
er
e
n
t
w
a
y
s
to
h
an
d
le
m
is
s
in
g
d
ata.
Do
in
g
m
o
r
e
r
esear
ch
,
th
e
tea
m
f
o
u
n
d
ac
tu
al
s
u
r
v
e
y
s
o
f
w
h
er
e
s
o
m
e
o
th
er
r
esear
ch
er
s
h
a
v
e
u
s
ed
o
n
e
o
f
th
e
f
i
v
e
m
eth
o
d
s
d
is
c
u
s
s
ed
p
r
ev
io
u
s
l
y
.
T
h
e
to
p
tw
o
s
u
r
v
e
y
s
w
ill
b
e
d
is
cu
s
s
ed
[
8
-
10
]
.
I
n
th
e
f
ir
s
t
s
u
r
v
e
y
,
t
h
er
e
ar
e
t
w
o
m
eth
o
d
s
co
m
p
ar
ed
w
h
e
n
u
s
in
g
to
h
elp
ed
u
ca
tio
n
al
r
esear
ch
.
T
h
e
t
w
o
m
et
h
o
d
s
t
h
at
w
er
e
co
m
p
ar
ed
w
er
e
s
tep
w
is
e
r
eg
r
es
s
io
n
w
h
ic
h
i
s
t
h
e
s
a
m
e
as
ig
n
o
r
in
g
t
h
e
m
is
s
i
n
g
d
ata
v
er
s
u
s
m
u
ltip
le
i
m
p
u
ta
tio
n
.
T
h
r
o
u
g
h
o
u
t
th
e
a
r
ticle,
it
talk
s
ab
o
u
t
h
o
w
in
r
eg
r
ess
io
n
m
o
d
elin
g
th
e
s
tep
w
is
e
r
eg
r
e
s
s
io
n
ca
n
b
e
d
an
g
er
o
u
s
in
m
is
s
i
n
g
f
iles
as
th
e
lo
s
s
o
f
i
n
f
o
r
m
atio
n
ca
n
b
e
cr
itical
an
d
th
e
an
al
y
s
t
m
a
y
n
o
t
e
v
en
n
o
tice
it
.
T
h
is
is
co
m
p
ar
ed
to
o
n
e
o
f
th
e
f
iv
e
m
et
h
o
d
s
r
esear
ch
ed
in
th
e
s
ec
tio
n
s
ab
o
v
e
.
I
t
ess
en
t
iall
y
r
e
m
o
v
e
s
t
h
e
w
h
o
le
en
tr
y
f
o
r
an
y
b
la
n
k
s
f
o
u
n
d
in
t
h
e
d
ata
a
n
d
th
e
ar
ticle
a
g
r
ee
s
th
at
it
ca
n
h
av
e
p
o
ten
tial
lo
s
s
o
f
s
u
b
s
ta
n
tial
a
m
o
u
n
ts
o
f
i
n
f
o
r
m
atio
n
w
h
i
le
also
in
cr
ea
s
i
n
g
t
h
e
b
ias
f
o
r
th
e
m
o
d
els
ap
p
lied
.
T
h
en
,
th
e
ar
ticle
d
is
cu
s
s
es
h
o
w
m
u
l
tip
le
i
m
p
u
tatio
n
is
n
o
t
y
et
b
ein
g
ac
ce
p
ted
b
y
ev
er
y
o
n
e
b
ec
au
s
e
it
in
v
o
l
v
es
ad
d
in
g
s
i
m
u
lated
d
ata
to
a
r
a
w
d
ata
s
et
an
d
p
eo
p
le
th
in
k
th
at
th
er
e
is
a
m
a
n
ip
u
la
tio
n
in
t
h
e
d
ata
an
d
th
er
ef
o
r
e
n
o
t
a
g
o
o
d
r
e
p
r
esen
tatio
n
o
f
th
e
o
r
ig
in
al
d
ata.
Yet,
i
m
p
u
tatio
n
d
o
es
th
e
o
p
p
o
s
ite
as
it
tr
ies
to
u
s
e
th
e
in
f
o
r
m
atio
n
a
v
ailab
le
to
s
i
m
u
l
ate
th
e
m
is
s
in
g
d
ata
to
m
i
n
i
m
i
ze
th
e
b
ias.
T
h
e
s
ec
o
n
d
s
u
r
v
e
y
r
ev
ie
w
ed
t
h
e
c
h
allen
g
e
s
o
f
m
i
s
s
i
n
g
d
ata
in
ap
p
l
y
in
g
it
to
m
ed
ical
r
ese
ar
ch
.
T
h
e
ar
ticle
f
u
r
th
er
d
is
c
u
s
s
ed
an
i
m
p
u
tat
io
n
m
eth
o
d
s
tat
in
g
t
h
at
m
an
y
p
eo
p
le
w
er
e
n
o
t
co
n
s
i
d
er
in
g
it
a
n
d
in
s
tead
w
er
e
c
o
n
s
id
er
in
g
ca
s
e
-
w
is
e
d
eletio
n
m
et
h
o
d
.
T
h
e
d
ata
w
a
s
a
clas
s
i
f
icatio
n
d
ataset
a
n
d
it
h
elp
ed
to
v
alid
ate
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2586
Mis
s
in
g
d
a
ta
h
a
n
d
lin
g
fo
r
ma
ch
in
e
lea
r
n
in
g
m
o
d
els
(
K
a
r
im
H.
E
r
ia
n
)
125
an
d
co
n
f
ir
m
o
n
e
o
f
th
e
f
i
v
e
m
et
h
o
d
s
d
is
cu
s
s
ed
ab
o
v
e
in
s
a
y
i
n
g
th
at
t
h
e
i
m
p
u
tatio
n
m
eth
o
d
is
n
o
t
g
o
o
d
en
o
u
g
h
f
o
r
h
an
d
li
n
g
th
e
m
is
s
in
g
d
ata.
I
t
d
is
cu
s
s
es
h
a
n
d
li
n
g
m
i
s
s
i
n
g
d
ata
at
r
an
d
o
m
v
er
s
u
s
n
o
t
at
r
an
d
o
m
s
tati
n
g
t
h
e
p
r
o
b
lem
s
w
it
h
alter
n
ati
v
e
to
d
ata
m
an
ip
u
latio
n
.
2
.
6
.
P
r
o
po
s
ed
a
pp
ro
a
ch
T
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
to
s
o
l
v
e
th
e
p
r
o
b
le
m
o
f
m
i
s
s
i
n
g
d
at
a
is
to
tak
e
i
n
to
co
n
s
id
er
atio
n
th
e
u
n
iq
u
e
v
alu
e
o
f
a
ce
ll
d
ir
e
ctly
,
i
n
s
te
ad
o
f
j
u
s
t
d
eletin
g
th
e
w
h
o
le
en
tr
y
.
I
n
s
tead
o
f
d
eleti
n
g
t
h
e
w
h
o
le
e
n
tr
y
w
h
en
f
i
n
d
in
g
a
b
lan
k
f
ield
,
th
e
tea
m
o
n
l
y
d
elete
d
t
h
at
o
n
e
s
p
ec
if
ic
ce
ll.
T
h
e
tea
m
co
m
b
i
n
ed
m
u
ltip
le
1
-
f
ea
t
u
r
e
m
o
d
el
s
to
cr
ea
te
a
n
e
w
m
o
d
el
.
I
n
th
is
m
an
n
er
,
th
e
m
o
d
el
cr
ea
t
ed
is
f
ir
s
t
tr
ai
n
ed
b
y
ea
c
h
f
ea
tu
r
e
alo
n
e,
th
e
n
w
ei
g
h
ts
w
er
e
co
m
b
in
ed
to
g
et
h
er
.
I
n
th
e
‘
Me
th
o
d
s
’
s
ec
tio
n
,
t
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
is
f
u
r
th
er
ed
d
is
cu
s
s
ed
w
it
h
th
e
tr
ai
n
in
g
o
f
th
e
2
2
0
d
if
f
er
e
n
t
m
o
d
el
s
an
d
f
ea
t
u
r
es.
Ha
v
in
g
all
d
i
f
f
er
e
n
t
w
ei
g
h
ts
f
r
o
m
th
e
d
if
f
er
e
n
t
m
o
d
els,
th
e
y
w
il
l
b
e
co
m
b
in
ed
to
g
et
h
er
to
h
av
e
o
n
e
b
ig
m
o
d
el
w
it
h
o
u
t
j
eo
p
ar
d
izin
g
th
e
in
teg
r
i
t
y
o
f
th
e
f
u
ll
m
i
s
s
i
n
g
en
tr
ies,
y
e
t sti
l
l b
ein
g
ab
le
to
g
ath
er
m
o
r
e
ac
cu
r
ate
d
ata
f
o
r
th
e
f
ea
t
u
r
es.
3.
RE
S
E
ARCH
M
E
T
H
O
DS A
ND
I
M
P
L
E
M
E
NT
AT
I
O
N
3
.
1
.
P
re
-
pro
ce
s
s
ing
t
he
da
t
a
I
n
th
is
p
h
ase,
t
h
e
tea
m
h
ad
to
u
n
d
er
s
tan
d
t
h
e
a
v
ailab
le
d
ata
,
h
o
w
it
is
l
in
k
ed
to
g
et
h
er
,
a
n
d
h
o
w
to
f
i
n
d
a
tar
g
et
f
o
r
an
y
e
n
tr
y
in
th
e
C
SV
f
ile
s
.
I
t
r
eq
u
ir
ed
a
l
o
t
o
f
in
ten
s
iv
e
r
ea
d
in
g
an
d
r
e
s
ea
r
ch
i
n
g
p
r
e
v
io
u
s
i
m
p
le
m
en
ta
tio
n
s
to
s
ee
w
h
a
t
t
ests
o
t
h
er
s
h
a
v
e
i
m
p
le
m
e
n
ted
an
d
w
h
y
it
w
o
r
k
ed
o
r
d
id
n
o
t
w
o
r
k
at
a
ce
r
tai
n
p
r
ed
ictio
n
ac
cu
r
ac
y
[
11
-
13
]
.
T
h
en
,
it
w
as
n
o
ticed
th
at
t
h
e
d
ata
in
clu
d
ed
lo
ts
o
f
s
tr
in
g
d
a
tat
y
p
e
v
ar
iab
les.
Fo
r
ex
a
m
p
le,
‘
Do
es
th
e
ap
p
lica
n
t
h
av
e
a
ca
r
?
’
I
t
w
o
u
ld
eit
h
er
b
e
a
‘
y
es
’
s
tr
i
n
g
v
ar
iab
le
o
r
a
‘
n
o
’
s
tr
in
g
v
ar
iab
le.
Strin
g
d
atat
y
p
es
ca
n
b
e
v
er
y
c
o
m
p
le
x
to
p
r
o
ce
s
s
,
s
o
a
s
er
ies
o
f
d
ata
d
ictio
n
ar
ies
w
er
e
i
m
p
l
e
m
en
ted
to
m
ap
all
v
alu
e
s
f
r
o
m
ca
te
g
o
r
ical
v
alu
e
s
to
n
u
m
er
ical
v
al
u
es.
Fo
r
th
e
c
o
n
v
er
ted
ce
lls
,
if
th
e
ce
ll
w
a
s
e
m
p
t
y
a
v
al
u
e
w
a
s
ass
i
g
n
ed
to
a
n
i
n
te
g
er
v
ar
iab
le
o
f
‘
9
9
9
9
9
9
9
9
’
.
A
l
s
o
,
th
e
tar
g
et
v
al
u
e
w
as
m
ap
p
ed
to
an
i
n
teg
er
v
ar
iab
le
o
f
1
id
en
ti
f
y
i
n
g
t
h
at
it
w
a
s
f
i
n
e
to
len
d
th
e
ap
p
lican
t
m
o
n
e
y
a
n
d
an
in
te
g
er
v
ar
iab
le
o
f
0
to
i
d
en
tify
a
h
ig
h
r
is
k
ap
p
lican
t.
T
h
en
,
th
e
lo
g
i
s
tic
r
eg
r
ess
io
n
m
o
d
el
w
as
tr
ai
n
ed
o
n
8
0
%
o
f
th
e
d
ata
an
d
test
e
d
u
s
in
g
2
0
%
o
f
t
h
e
d
ata.
T
h
e
r
esu
lts
g
a
v
e
an
ac
c
ep
tab
le
ac
cu
r
ac
y
p
er
ce
n
ta
g
e
o
f
9
1
%.
Fig
u
r
e
1
is
a
s
a
m
p
le
o
f
th
e
d
ictio
n
ar
ies
i
m
p
le
m
en
ted
.
Fig
u
r
e
1
.
Sa
m
p
le
d
ictio
n
ar
ies
cr
ea
ted
3
.
2
.
Understa
nd
ing
ea
ch
f
ea
t
ure
i
m
pa
ct
o
n t
a
rg
et
I
n
o
r
d
er
t
o
u
n
d
er
s
tan
d
ea
ch
f
e
atu
r
e
i
m
p
ac
t
o
n
t
h
e
tar
g
et
v
a
l
u
e
(
w
h
ic
h
f
lo
w
c
h
ar
t
is
s
h
o
w
n
b
y
Fi
g
u
r
e
2)
,
th
e
tea
m
h
ad
to
ca
lcu
la
te
t
h
e
ac
cu
r
ac
y
o
f
ea
ch
f
ea
t
u
r
e
alo
n
e.
T
h
is
w
a
s
d
o
n
e
as
p
ar
t
o
f
d
ata
u
n
d
er
s
ta
n
d
in
g
.
I
t
w
a
s
also
d
o
n
e
at
f
ir
s
t
w
h
en
tr
y
in
g
to
c
h
o
o
s
e
s
p
ec
i
f
ic
f
ea
t
u
r
es,
n
o
t
to
u
s
e
all
o
f
t
h
e
m
.
I
n
th
is
ca
s
e,
t
h
e
tea
m
i
m
p
le
m
en
ted
2
2
0
m
o
d
els
f
o
r
all
f
ea
tu
r
es,
ca
lcu
lated
th
e
p
er
ce
n
tag
e
an
d
s
a
v
ed
th
e
m
.
Ho
w
e
v
er
,
th
e
f
ir
s
t
1
2
0
f
ea
t
u
r
es
ac
cu
r
ac
ies
ar
e
r
ep
r
es
en
ted
in
Fig
u
r
e
3
,
i
t
w
as
n
o
t
iced
th
at
s
o
m
e
o
f
th
e
f
ea
t
u
r
es
ar
e
h
av
e
a
v
er
y
m
i
n
i
m
al
i
m
p
ac
t
o
f
7
%
p
r
ed
ictio
n
w
h
ile
o
t
h
er
s
h
ad
m
a
x
i
m
u
m
i
m
p
ac
t
s
o
f
u
p
to
9
0
%
ac
cu
r
ac
y
.
T
h
e
team
al
s
o
n
o
ticed
a
lo
t o
f
b
lan
k
s
w
er
e
p
r
esen
t i
n
s
o
m
e
f
ea
t
u
r
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2586
I
A
E
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
10
,
No
.
2
,
J
u
n
e
2
0
2
1
:
1
2
3
–
132
126
Fig
u
r
e
2
.
Featu
r
es i
m
p
ac
t o
n
t
ar
g
et
f
lo
w
c
h
ar
t
Fig
u
r
e
3
.
I
m
a
g
e
r
ep
r
esen
ti
n
g
e
ac
h
f
ea
t
u
r
e
i
m
p
ac
t o
n
tar
g
et
p
r
ed
ictio
n
u
s
in
g
lo
g
i
s
tic
r
eg
r
e
s
s
i
o
n
3
.
3
.
Rem
o
v
ing
bla
n
ks
a
nd
c
hec
k
ing
t
he
a
cc
ura
cy
T
h
e
team
d
ec
id
ed
to
p
r
ep
ar
e
th
e
d
ata
m
o
r
e
to
h
av
e
it
in
a
b
etter
s
h
ap
e
an
d
b
etter
u
s
ag
e.
T
h
u
s
,
to
r
e
m
o
v
e
b
lan
k
s
an
d
s
tar
t
co
m
p
ar
in
g
f
ea
t
u
r
es
f
o
r
th
e
f
ea
t
u
r
e
s
elec
tio
n
p
ar
t.
T
h
e
team
f
o
u
n
d
th
at
b
lan
k
s
ar
e
h
an
d
led
m
ai
n
l
y
i
n
o
n
e
o
f
th
e
3
m
eth
o
d
s
.
E
ith
er
b
y
r
e
m
o
v
i
n
g
e
n
tr
ies,
o
r
b
y
f
illi
n
g
in
th
e
b
lan
k
s
b
y
d
ef
a
u
l
t
v
alu
e
s
o
r
ze
r
o
s
.
T
h
e
tea
m
tr
ied
r
em
o
v
i
n
g
t
h
e
w
h
o
le
e
n
tr
ies,
b
u
t
t
h
at
ca
u
s
e
t
h
e
d
ata
s
ize
to
s
h
r
i
n
k
f
r
o
m
3
0
7
,
5
1
1
en
tr
ies
d
o
w
n
to
8
,
6
0
3
en
tr
ies.
Als
o
,
th
e
p
r
ed
ictio
n
ac
cu
r
ac
y
b
ec
a
m
e
m
u
c
h
lo
w
er
an
d
t
h
e
f
ea
t
u
r
e
ef
f
ec
t
o
n
th
e
tar
g
et
p
r
ed
ictio
n
d
ec
r
ea
s
ed
s
ig
n
i
f
ica
n
tl
y
.
T
h
is
w
as
m
ain
l
y
th
e
m
ain
m
o
t
iv
at
io
n
to
s
w
itc
h
th
e
f
o
cu
s
o
n
h
o
w
to
h
an
d
le
m
is
s
in
g
d
ata.
Fig
u
r
e
4
illu
s
tr
ates
th
e
i
m
p
ac
t
o
f
ea
ch
f
ea
t
u
r
e
o
n
tar
g
et
p
r
ed
ictio
n
,
w
h
ile
Fi
g
u
r
es 5
-
6
s
h
o
w
s
t
h
e
p
r
o
ce
s
s
f
lo
w
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2586
Mis
s
in
g
d
a
ta
h
a
n
d
lin
g
fo
r
ma
ch
in
e
lea
r
n
in
g
m
o
d
els
(
K
a
r
im
H.
E
r
ia
n
)
127
Fig
u
r
e
4
.
Featu
r
e
i
m
p
ac
t
on
ta
r
g
et
p
r
ed
ictio
n
u
s
i
n
g
lo
g
is
tic
r
eg
r
ess
io
n
af
ter
r
e
m
o
v
i
n
g
e
n
tr
i
es
w
ith
m
i
s
s
i
n
g
d
at
a
Fig
u
r
e
5
.
C
o
m
b
in
ed
f
ea
tu
r
e
s
ac
cu
r
ac
y
ca
lc
u
latio
n
f
lo
w
c
h
ar
t
Fig
u
r
e
6
.
B
lan
k
s
s
p
ec
ial
tr
ea
tm
en
t f
lo
w
c
h
ar
t
3
.
4
.
H
a
nd
lin
g
m
is
s
ing
da
t
a
T
h
e
im
p
le
m
e
n
ted
ap
p
r
o
ac
h
is
ab
o
u
t
h
an
d
li
n
g
m
is
s
i
n
g
d
ata
in
a
n
e
w
m
ac
h
in
e
lear
n
in
g
m
o
d
el.
A
f
ter
r
e
m
o
v
i
n
g
all
e
n
tr
ies
w
it
h
m
is
s
in
g
d
ata,
t
h
e
ac
c
u
r
ac
y
d
ec
li
n
ed
en
o
r
m
o
u
s
l
y
.
Fo
r
th
i
s
r
ea
s
o
n
,
t
h
e
tea
m
d
ec
id
ed
to
h
av
e
a
n
o
th
er
m
o
d
el
th
at
is
a
co
m
b
i
n
atio
n
o
f
m
u
ltip
le
1
-
f
ea
tu
r
e
m
o
d
els.
I
n
ea
ch
m
o
d
el
,
th
e
tea
m
h
a
n
d
led
th
e
m
i
s
s
i
n
g
d
ata
in
o
n
e
o
f
t
h
e
alr
ea
d
y
k
n
o
w
n
m
et
h
o
d
s
w
h
i
ch
in
cl
u
d
ed
ap
p
ly
i
n
g
th
e
f
ea
t
u
r
e’
s
d
ef
au
lt
v
al
u
e,
g
etti
n
g
t
h
e
av
er
a
g
e
if
t
h
e
f
ea
tu
r
e
is
a
n
u
m
er
ical
d
ata,
o
r
ju
s
t
r
e
m
o
v
i
n
g
t
h
e
m
is
s
in
g
ce
l
ls
.
T
h
e
m
o
d
el
w
as
ap
p
lied
to
ea
ch
f
ea
t
u
r
e
alo
n
e
i
n
o
r
d
er
to
f
i
n
d
th
e
b
e
s
t
ap
p
r
o
ac
h
f
o
r
ea
c
h
o
n
e.
T
h
e
n
e
x
t
s
tep
th
e
tea
m
to
o
k
w
as
to
tr
ain
2
2
0
d
if
f
er
en
t
m
o
d
els
f
o
r
th
e
2
2
0
f
ea
tu
r
es.
T
h
en
,
h
av
i
n
g
th
e
d
if
f
er
en
t
w
ei
g
h
t
s
f
r
o
m
th
e
d
if
f
er
en
t
m
o
d
el
s
,
t
h
e
tea
m
co
m
b
i
n
ed
t
h
e
m
to
g
et
h
er
to
h
a
v
e
o
n
e
m
o
d
el
th
at
i
s
n
o
t
r
u
i
n
ed
b
y
m
is
s
in
g
d
ata
an
d
at
th
e
s
a
m
e
ti
m
e,
g
e
ttin
g
ac
c
u
r
ate
d
ata
f
o
r
all
f
ea
t
u
r
es.
Ho
w
e
v
er
,
th
i
s
w
a
s
n
o
t
en
o
u
g
h
b
ec
au
s
e
if
th
e
f
ea
t
u
r
e
u
s
ed
alo
n
e
to
tr
ain
th
e
d
ata
is
h
ar
d
l
y
r
elate
d
to
th
e
t
ar
g
et,
it
m
a
y
a
f
f
ec
t
t
h
e
o
v
er
all
m
o
d
el
[
8
]
(
s
ee
Fig
u
r
e
7
)
.
So
f
u
r
t
h
er
m
a
th
e
m
atica
l
s
o
lu
t
io
n
s
w
er
e
ap
p
lied
(
as illu
s
tr
ated
b
y
Fi
g
u
r
e
8
)
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2586
I
A
E
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
10
,
No
.
2
,
J
u
n
e
2
0
2
1
:
1
2
3
–
132
128
‑
Mu
ltip
l
y
th
e
n
e
w
w
ei
g
h
ts
b
y
o
ld
w
ei
g
h
t
s
f
r
o
m
th
e
m
o
d
el
u
s
e
d
to
tr
ain
th
e
2
2
0
f
ea
tu
r
es to
g
e
th
er
.
‑
Mu
ltip
l
y
th
e
n
e
w
w
ei
g
h
ts
b
y
t
h
e
test
i
n
g
ac
cu
r
ac
y
f
r
o
m
ea
c
h
s
ep
ar
ated
m
o
d
el
b
ef
o
r
e
r
e
m
o
v
in
g
b
la
n
k
s
.
=
0
.
0
+
0
=
1
.
+
1
+
1
=
2
.
2
+
2
=
3
.
3
+
3
=
4
.
+
4
+
4
=
5
.
5
+
5
=
6
.
6
+
6
=
7
.
+
7
+
7
if
:
w
=
[
0
1
2
3
4
5
6
7
]
a
n
d
:
b
=
[
0
1
2
3
4
5
6
7
]
th
en
=
.
+
=
0
.
0
+
0
+
1
.
+
1
+
1
+
2
.
2
+
2
+
3
.
3
+
3
+
4
.
+
4
+
4
+
5
.
5
+
5
+
6
.
6
+
6
+
7
.
+
7
+
7
B
u
t th
i
s
is
eq
u
al
to
8
T
n
o
t T
.
Fig
u
r
e
7
.
Old
m
e
th
o
d
o
f
h
a
n
d
l
in
g
m
i
s
s
i
n
g
d
ata,
all
w
h
ite
s
q
u
ar
es a
r
e
p
r
esen
tin
g
m
is
s
in
g
d
ata
Fig
u
r
e
8
.
P
r
o
p
o
s
ed
m
eth
o
d
o
f
h
an
d
li
n
g
m
i
s
s
i
n
g
d
ata
(
w
h
ite
s
q
u
ar
es =
m
i
s
s
i
n
g
d
ata)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2586
Mis
s
in
g
d
a
ta
h
a
n
d
lin
g
fo
r
ma
ch
in
e
lea
r
n
in
g
m
o
d
els
(
K
a
r
im
H.
E
r
ia
n
)
129
a.
Me
th
o
d
I
:
Div
i
d
e
it
b
y
th
e
n
u
m
b
er
o
f
f
ea
t
u
r
es.
I
n
t
h
is
e
x
a
m
p
le,
it
s
h
o
u
ld
b
e
d
iv
id
ed
b
y
8
f
ea
tu
r
es,
b
u
t
i
n
th
e
p
r
o
j
ec
t,
it w
as d
i
v
id
ed
b
y
2
2
0
f
ea
tu
r
es.
=
.
+
=
(
0
0
+
0
+
1
1
+
1
+
2
2
+
2
+
3
3
+
3
+
4
4
+
4
+
5
5
+
5
+
6
6
+
6
+
7
7
+
7
)
/
8
=
[
0
1
2
3
4
5
6
7
]
/
8
a
n
d
:
=
[
0
1
2
3
4
5
6
7
]
/
8
b.
Me
th
o
d
I
I
:
W
h
en
lin
ea
r
r
eg
r
e
s
s
io
n
w
a
s
f
ir
s
t
u
s
ed
,
th
e
w
ei
g
h
t
v
ec
to
r
w
as
d
er
iv
ed
,
(
ass
u
m
e
it
as
’
v
’
i
n
s
tead
o
f
’
w
’
n
o
t to
co
n
f
l
ict
it
w
it
h
t
h
e
’
w
’
t
h
at
h
a
s
alr
ea
d
y
b
ee
n
d
is
cu
s
s
ed
)
.
L
et
v
=
[
v
0
v
1
v
2
v
3
v
4
v
5
v
6
v
7
]
.
Sin
ce
it
is
k
n
o
w
n
th
a
t
th
e
w
eig
h
t
s
u
m
i
s
al
w
a
y
s
eq
u
al
to
1
,
th
en
an
eq
u
atio
n
ca
n
b
e
estab
lis
h
ed
s
u
ch
t
h
at
v
0
+
v
1
+
v
2
+
v
3
+
v
4
+
v
5
+
v
6
+
v
7
=
1
.
I
n
th
i
s
ca
s
e,
ea
ch
elem
e
n
t
o
f
w
w
a
s
m
u
l
tip
lied
b
y
t
h
e
co
r
r
esp
o
n
d
in
g
ele
m
en
t
o
f
v
.
T
h
e
h
i
g
h
er
i
m
p
ac
t
w
ei
g
h
t
in
th
e
w
v
e
cto
r
is
en
h
an
ce
d
b
y
t
h
e
w
e
i
g
h
t
f
r
o
m
t
h
e
i
n
i
t
i
a
l
l
i
n
e
a
r
r
e
g
r
e
s
s
i
o
n
w
e
i
g
h
t
s
v
.
T
h
e
r
e
f
o
r
e
,
w
=
[
w
0
.v
0
,w
1
.v
1
,w
2
.v
2
,w
3
.v
3
,w
4
.v
4
,w
5
.v
5
,w
6
.v
6
,w
7
.v
7
]
.
3
.
4
.
Cre
a
t
ing
2
2
0
m
o
dels
T
o
im
p
le
m
e
n
t
t
h
is
ap
p
r
o
ac
h
,
th
e
tea
m
n
ee
d
ed
to
b
u
ild
a
m
ac
h
in
e
lear
n
i
n
g
m
o
d
el
f
o
r
ea
ch
f
ea
t
u
r
e
alo
n
e.
T
h
e
tea
m
d
ec
id
ed
to
u
s
e
th
e
l
in
ea
r
r
e
g
r
ess
io
n
m
o
d
el
f
o
r
t
w
o
m
ai
n
r
ea
s
o
n
s
.
T
h
e
f
ir
s
t
m
ai
n
r
ea
s
o
n
t
h
at
a
lin
ea
r
r
eg
r
e
s
s
io
n
m
o
d
el
w
as
u
s
ed
w
as
b
ec
a
u
s
e
i
t
w
a
s
clea
n
e
r
to
b
u
ild
t
h
e
w
ei
g
h
ti
n
g
v
ec
to
r
f
r
o
m
t
h
e
s
ep
ar
ate
w
ei
g
h
ts
[
14
]
.
T
h
e
s
ec
o
n
d
r
e
aso
n
th
at
a
lin
ea
r
r
eg
r
ess
io
n
m
o
d
el
w
a
s
u
s
ed
w
a
s
b
ec
au
s
e
it
h
ad
a
h
ig
h
er
ac
cu
r
ac
y
i
n
n
o
r
m
al
ca
s
es
f
o
r
th
is
d
ataset
t
h
a
n
th
e
lo
g
i
s
tic
r
eg
r
ess
io
n
m
o
d
el
w
h
ic
h
co
m
p
ar
ed
9
3
%
v
s
9
1
%,
r
esp
ec
tiv
el
y
.
T
o
im
p
le
m
en
t
th
e
m
o
d
els,
t
h
e
tea
m
s
et
u
p
th
e
‘
X
’
m
atr
i
x
f
ir
s
t.
T
h
en
,
2
2
0
‘
X
’
v
ec
to
r
s
w
e
r
e
b
u
ilt
in
ea
ch
o
n
e.
T
h
er
ea
f
ter
,
all
th
e
b
lan
k
d
ata
w
er
e
r
e
m
o
v
ed
f
r
o
m
all
th
e
g
at
h
er
ed
C
SV
f
ile
s
.
Af
ter
th
i
s
s
et
u
p
,
th
e
tea
m
co
n
ti
n
u
ed
to
g
ath
er
t
h
e
co
r
r
esp
o
n
d
in
g
tar
g
et
v
ec
to
r
f
o
r
ea
ch
in
p
u
t
d
ata
v
ec
to
r
.
T
h
is
w
a
s
d
o
n
e
b
ec
au
s
e
w
h
e
n
th
e
b
lan
k
d
ata
is
r
e
m
o
v
ed
f
r
o
m
o
n
l
y
o
n
e
co
l
u
m
n
(
n
o
t
all
co
lu
m
n
s
to
g
e
th
er
)
,
th
e
‘
T
’
v
ec
to
r
w
as
n
ee
d
ed
to
m
a
tch
t
h
e
d
ata
in
p
u
t.
I
n
th
is
p
r
o
ce
s
s
,
th
e
ch
alle
n
g
e
w
a
s
th
at
n
o
t a
ll
C
S
V
f
ile
s
h
ad
t
h
e
tar
g
et
co
lu
m
n
.
A
lt
h
o
u
g
h
,
it
w
as
o
n
l
y
t
h
e
“
ap
p
licatio
n
tr
ain
.
csv
”
f
i
le,
th
e
r
est
o
f
th
e
f
iles
h
ad
an
attr
ib
u
te
th
at
w
a
s
ca
lled
“
Ke
y
I
D
C
u
r
r
e
n
t”
th
at
w
a
s
p
r
esen
t
in
s
o
m
e
o
f
th
e
C
SV
f
ile
s
in
cl
u
d
in
g
in
th
e
“
ap
p
licatio
n
tr
ain
.
cs
v
”
f
ile.
T
h
e
t
ea
m
u
s
ed
t
h
e
I
Ds
i
n
th
e
s
e
C
SV
f
iles
to
b
e
ab
le
to
lin
k
t
h
e
tar
g
e
t
to
th
e
i
n
p
u
t
d
ata
f
r
o
m
th
e
o
t
h
er
C
SV
f
iles
.
T
h
en
,
it
w
as
n
o
ti
ce
d
th
at
s
o
m
e
o
f
th
e
C
SV
f
i
les
d
id
n
o
t
h
av
e
th
e
“
Ke
y
I
D
C
u
r
r
en
t”
attr
ib
u
te
p
r
esen
t,
b
u
t
th
er
e
w
as
an
o
t
h
er
I
D
th
at
ex
is
ted
th
at
co
u
ld
b
e
u
s
ed
w
ith
t
h
e
r
est
o
f
th
e
f
iles
b
ec
au
s
e
it
lin
k
ed
u
p
ea
ch
o
f
th
o
s
e
I
Ds
n
icel
y
f
o
r
a
ll
o
f
t
h
e
f
ile
s
.
I
n
t
h
is
m
a
n
n
er
,
th
e
tea
m
u
s
ed
th
e
s
ec
o
n
d
ar
y
I
D
f
o
u
n
d
to
g
et
th
e
“
T
”
v
ec
to
r
f
o
r
all
i
n
p
u
t d
ata
[
15
].
3
.
4
.
Cre
a
t
ing
new
m
o
del
No
w
,
h
a
v
in
g
th
e
2
2
0
’
X’
v
ec
t
o
r
s
an
d
th
e
2
2
0
’
T
’
v
ec
to
r
s
,
th
e
tea
m
d
iv
id
ed
th
e
m
i
n
to
8
0
%
tr
ain
in
g
d
ata
an
d
2
0
%
test
in
g
d
ata.
T
h
e
tea
m
tr
ain
ed
all
m
o
d
els
an
d
g
o
t
2
2
0
’
w
’
an
d
2
2
0
b
ias
v
alu
es.
Me
th
o
d
I
an
d
Me
th
o
d
I
I
w
er
e
i
m
p
le
m
e
n
ted
to
co
m
b
i
n
e
an
d
cr
ea
te
th
e
’
w
’
v
ec
to
r
f
o
r
th
e
n
e
w
m
o
d
el.
Fi
g
u
r
e
9
is
a
f
lo
w
c
h
ar
t
d
escr
ib
in
g
th
e
n
e
w
ap
p
r
o
ac
h
.
C
o
lo
r
lig
h
t
o
r
an
g
e
r
ep
r
esen
ts
th
e
n
o
v
el
id
ea
.
T
h
e
m
a
in
f
lo
w
ch
ar
t
co
n
tai
n
s
r
ec
tan
g
le
s
a
n
d
b
o
x
es.
T
h
e
b
o
x
es
ar
e
ex
p
lai
n
ed
in
m
o
r
e
d
etai
l
in
o
t
h
er
f
o
llo
w
in
g
f
lo
w
ch
ar
t
s
w
it
h
s
i
m
ilar
co
lo
r
o
f
th
e
o
r
ig
i
n
al
b
o
x
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
4
.
1
.
Cre
a
t
ing
t
esting
da
t
a
s
et
Fo
r
test
i
n
g
,
t
h
e
tea
m
co
u
ld
n
o
t
j
u
s
t
u
s
e
ea
c
h
f
ile
s
ep
ar
atel
y
li
k
e
it
w
as
d
o
n
e
f
o
r
t
h
e
tr
a
in
i
n
g
p
ar
t.
T
h
er
ef
o
r
e,
th
e
team
m
ap
p
ed
all
th
e
d
ata
f
r
o
m
all
f
ile
s
u
s
i
n
g
th
e
I
Ds
f
o
u
n
d
in
t
h
e
C
SV
f
iles
.
T
h
is
s
tep
to
o
k
w
a
y
m
o
r
e
ti
m
e
t
h
an
e
x
p
ec
ted
.
T
h
e
tea
m
s
ea
r
c
h
ed
f
o
r
a
m
e
th
o
d
to
d
o
th
is
p
r
o
ce
s
s
o
f
m
a
p
p
in
g
w
i
th
o
u
t
’
f
o
r
’
lo
o
p
s
to
r
ed
u
ce
tim
e
co
m
p
lex
i
t
y
,
esp
ec
iall
y
w
it
h
th
e
lar
g
e
d
ataset
th
at
w
as
av
ailab
le,
b
u
t
u
n
f
o
r
tu
n
atel
y
,
th
er
e
w
a
s
n
o
ea
s
y
m
et
h
o
d
f
o
u
n
d
.
T
h
e
m
eth
o
d
t
h
at
w
as
u
s
ed
to
co
m
b
i
n
e
t
h
e
d
ata
is
th
e
f
o
llo
win
g
:
Firs
t,
th
e
tea
m
in
itial
ized
a
n
e
w
‘
X
’
ar
r
a
y
a
s
a
s
et
o
f
ze
r
o
s
ar
r
a
y
w
i
th
t
h
e
s
i
ze
o
f
3
0
7
5
1
1
r
o
w
s
a
n
d
2
2
0
co
lu
m
n
s
.
T
h
en
,
u
s
i
n
g
th
e
“
ap
p
licatio
n
tr
ai
n
.
cs
v
”
f
ile
as
th
e
m
ai
n
f
ile,
th
e
tea
m
lo
o
p
ed
o
v
er
all
I
Ds
in
t
h
e
f
ile
an
d
s
ea
r
ch
ed
f
o
r
th
e
m
atc
h
ed
I
D
i
n
t
h
e
“
C
r
ed
it
C
a
r
d
B
alan
ce
.
csv
”
f
ile.
I
n
t
h
e
ca
s
e
th
a
t
a
m
atc
h
w
a
s
f
o
u
n
d
,
th
e
tea
m
ap
p
en
d
ed
th
e
v
alu
e
s
f
r
o
m
t
h
e
“
C
r
ed
it
C
ar
d
B
alan
ce
.
csv
”
f
ile
to
th
e
‘
X
’
ar
r
a
y
w
h
ic
h
alr
ea
d
y
h
ad
th
e
v
alu
e
s
f
r
o
m
t
h
e
“
ap
p
licatio
n
tr
ai
n
.
cs
v
”
f
ile.
T
h
e
tea
m
r
ep
ea
ted
th
is
p
r
o
ce
s
s
f
o
r
all
th
e
C
SV
f
iles
s
o
th
at
th
e
‘
X
’
m
a
tr
ix
w
a
s
av
ailab
le.
T
h
e
d
ata
w
as
t
h
en
s
ep
ar
ated
in
to
th
e
tr
ain
i
n
g
s
et
an
d
test
in
g
s
e
t,
s
u
c
h
t
h
at
th
e
test
i
n
g
d
ataset
w
a
s
n
o
t u
s
ed
b
ef
o
r
e
in
t
h
e
tr
ai
n
in
g
[
16
].
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2586
I
A
E
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
10
,
No
.
2
,
J
u
n
e
2
0
2
1
:
1
2
3
–
132
130
Fig
u
r
e
9
.
A
p
p
r
o
ac
h
m
ai
n
f
lo
w
ch
ar
t
4
.
2
.
Resul
t
s
f
o
r
t
esting
t
he
n
ew
m
o
de
l
Hav
in
g
th
e
test
in
g
d
atas
et
est
ab
l
is
h
e
d
an
d
th
e
‘
w
’
r
ea
d
y
f
o
r
th
e
tw
o
m
eth
o
d
s
,
t
esti
n
g
t
o
o
k
p
l
ac
e
.
I
t
w
as
f
o
u
n
d
th
at
th
e
a
cc
u
r
ac
y
f
o
r
b
o
th
m
eth
o
d
s
in
c
r
e
ase
d
t
o
9
8
%
.
T
h
is
is
a
5
%
in
c
r
e
ase
in
p
r
e
d
ic
ti
o
n
ac
cu
r
ac
y
f
r
o
m
th
e
o
l
d
m
eth
o
d
o
l
o
g
y
w
h
ich
h
ad
an
o
r
ig
in
al
ac
cu
r
ac
y
o
f
9
3
%.
T
h
e
t
ea
m
h
a
d
t
o
r
ev
i
ew
to
f
in
d
o
u
t
w
h
y
b
o
th
m
eth
o
d
s
r
esu
lt
ed
in
9
8
%
p
r
e
d
ic
ti
o
n
a
cc
u
r
ac
y
.
T
h
e
t
ea
m
f
o
u
n
d
th
at
th
e
d
if
f
e
r
en
c
e
in
th
e
tw
o
m
eth
o
d
s
d
i
d
n
o
t
af
f
ec
t
th
e
f
in
a
l
r
esu
lt
b
y
a
la
r
g
e
m
ar
g
in
b
e
ca
u
s
e
‘
Y
’
is
n
o
r
m
aliz
e
d
a
t
th
e
en
d
.
I
f
th
e
v
a
lu
e
en
d
e
d
u
p
b
e
in
g
g
r
ea
te
r
th
an
0
.
5
,
th
en
it
w
as r
o
u
n
d
ed
u
p
t
o
1
,
an
d
if
th
e
v
alu
e
w
as less
th
an
0
.
5
,
th
en
it w
as r
o
u
n
d
e
d
d
o
w
n
to
0
.
4
.
3
.
Dis
cus
s
io
n
4
.
3
.
1
.
Appl
ica
t
io
n
T
h
is
r
esear
ch
h
as
b
ee
n
d
o
n
e
in
p
ar
allel
to
an
o
t
h
er
r
esear
ch
ab
o
u
t
a
u
to
n
o
m
o
u
s
co
n
tr
o
l
o
f
an
al
l
-
ter
r
ain
v
e
h
icle
(
A
T
V)
in
t
h
e
w
o
o
d
s
.
T
h
e
r
esear
ch
w
as
d
i
v
id
ed
in
to
p
h
ase
s
.
T
h
e
f
ir
s
t
p
h
ase
[
1
7
-
19
]
w
a
s
to
co
n
tr
o
l
th
e
A
T
V
u
s
in
g
al
g
o
r
ith
m
s
.
T
h
e
s
ec
o
n
d
p
h
ase
w
as
t
o
g
ath
er
d
ata
f
r
o
m
t
h
e
A
T
V
w
h
ile
m
o
v
i
n
g
[
20
]
an
d
u
s
e
t
h
e
d
ata
to
i
m
p
le
m
en
t
a
r
atin
g
m
ec
h
an
i
s
m
to
id
en
tify
d
if
f
er
e
n
t
d
r
iv
i
n
g
b
eh
a
v
io
r
s
.
T
h
e
th
ir
d
p
h
ase
i
s
to
u
s
e
th
e
d
ata
g
ath
er
ed
w
ith
th
e
m
et
h
o
d
in
v
e
n
ted
in
t
h
is
p
ap
er
in
o
r
d
er
to
co
n
tr
o
l
th
e
A
T
V
u
s
in
g
m
ac
h
i
n
e
lear
n
in
g
tec
h
n
iq
u
es.
4
.
3
.
2
.
L
ess
o
ns
l
ea
rne
d
T
h
r
o
u
g
h
th
is
p
r
o
j
ec
t,
th
e
team
w
as
ab
le
to
lear
n
a
v
ar
iet
y
o
f
tech
n
ical
a
n
d
s
o
f
t
s
k
ill
s
.
So
m
e
o
f
th
e
tech
n
ical
s
k
i
lls
lear
n
ed
th
r
o
u
g
h
th
is
p
r
o
j
ec
t
w
er
e
th
e
ex
p
lo
r
atio
n
o
f
th
e
lin
ea
r
r
eg
r
ess
io
n
m
o
d
el
in
co
m
p
ar
is
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2586
Mis
s
in
g
d
a
ta
h
a
n
d
lin
g
fo
r
ma
ch
in
e
lea
r
n
in
g
m
o
d
els
(
K
a
r
im
H.
E
r
ia
n
)
131
to
th
e
lo
g
is
tic
r
eg
r
es
s
io
n
m
o
d
el
to
s
ee
w
h
ic
h
w
o
u
ld
b
est
f
it
th
e
d
ata
o
f
th
is
s
p
ec
if
ic
class
i
f
icatio
n
p
r
o
j
ec
t.
A
t
f
ir
s
t,
o
n
e
o
f
t
h
e
m
o
s
t
ch
alle
n
g
in
g
th
in
g
s
f
o
r
t
h
e
tea
m
w
a
s
ch
o
o
s
in
g
a
p
r
o
j
ec
t
w
h
ic
h
co
n
ta
i
n
ed
en
o
u
g
h
d
ata.
I
n
m
ac
h
in
e
lear
n
i
n
g
,
th
e
tea
m
le
ar
n
ed
th
at
i
t
is
v
er
y
i
m
p
o
r
tan
t
to
h
av
e
a
s
m
u
ch
r
a
w
d
ata
a
s
p
o
s
s
ib
le.
T
h
is
w
as
ess
e
n
tial
f
o
r
tr
ain
i
n
g
a
n
d
b
ein
g
ab
le
to
g
en
er
alize
b
etter
f
o
r
test
in
g
w
h
ic
h
ev
er
m
o
d
el
ch
o
s
en
.
An
o
th
er
s
k
ill
lear
n
ed
th
r
o
u
g
h
t
h
is
r
esear
c
h
p
r
o
j
ec
t
is
th
at
it
is
v
e
r
y
i
m
p
o
r
ta
n
t
to
p
r
e
-
p
r
o
ce
s
s
th
e
d
ata
ca
r
ef
u
ll
y
a
n
d
th
o
r
o
u
g
h
l
y
.
I
f
th
e
d
ata
is
n
o
t
p
r
e
-
p
r
o
ce
s
s
ed
p
r
o
p
er
ly
,
it
w
i
ll
b
e
tr
ain
ed
i
n
t
h
e
w
r
o
n
g
m
a
n
n
er
a
n
d
t
h
er
ef
o
r
e
g
iv
e
s
k
e
w
ed
r
esu
lt
s
.
T
h
er
ef
o
r
e,
p
r
e
-
p
r
o
ce
s
s
in
g
t
h
e
d
ata
is
a
s
i
m
p
o
r
tan
t
i
f
n
o
t
m
o
r
e
i
m
p
o
r
tan
t
t
h
a
n
tr
ai
n
i
n
g
it
an
d
test
i
n
g
i
t.
Oth
er
s
k
ills
t
h
at
w
er
e
lear
n
ed
th
r
o
u
g
h
o
u
t
t
h
is
p
r
o
j
ec
t
w
er
e
th
e
co
n
tin
u
o
u
s
i
m
p
le
m
en
tatio
n
o
f
p
r
o
g
r
am
m
i
n
g
in
P
y
th
o
n
lan
g
u
ag
e
an
d
u
s
i
n
g
its
lib
r
ar
ies
to
m
ak
e
u
s
e
o
f
all
t
h
e
r
eso
u
r
ce
s
av
ailab
le
to
an
al
y
ze
th
e
d
ata.
T
h
is
w
as
a
lea
r
n
i
n
g
cu
r
v
e,
b
u
t
it
h
elp
ed
i
n
t
h
e
lo
n
g
r
u
n
i
n
b
ei
n
g
ab
le
to
h
an
d
le
d
ata
f
r
o
m
m
u
ltip
le
C
SV
f
i
les
at
th
e
s
a
m
e
ti
m
e.
T
h
is
alo
n
g
w
i
th
h
a
n
d
li
n
g
P
y
th
o
n
’
s
clas
s
es,
m
et
h
o
d
s
,
an
d
v
ar
iab
les
b
et
w
ee
n
m
u
ltip
le
J
u
p
y
ter
No
teb
o
o
k
s
a
n
d
m
u
lt
ip
le
m
ac
h
i
n
es
w
er
e
t
h
in
g
s
t
h
at
h
a
d
to
b
e
lear
n
ed
in
o
r
d
er
f
o
r
th
e
tea
m
n
o
t
to
w
a
s
te
ti
m
e
d
o
in
g
th
in
g
s
m
a
n
u
a
ll
y
o
r
o
n
e
s
tep
at
a
tim
e.
Ot
h
er
m
i
n
o
r
less
o
n
s
w
er
e
lear
n
ed
s
u
c
h
t
h
at
it
h
elp
ed
r
ed
u
ce
th
e
f
ile
s
ize
an
d
co
m
p
u
tatio
n
ti
m
e.
E
x
a
m
p
les
in
c
lu
d
ed
th
i
n
g
s
s
u
ch
as
r
e
m
o
v
i
n
g
‘
p
r
in
t
’
co
m
m
a
n
d
s
i
f
th
e
y
w
er
e
n
o
t
n
ee
d
ed
,
r
e
d
u
cin
g
lo
o
p
s
,
r
ed
u
ci
n
g
t
h
e
n
u
m
b
er
o
f
ce
lls
i
n
a
n
o
teb
o
o
k
,
an
d
h
av
in
g
s
ep
ar
ate
p
r
o
j
ec
t c
lass
es in
to
m
u
ltip
le
n
o
teb
o
o
k
s
.
A
l
l o
f
t
h
is
m
ad
e
th
e
p
r
e
-
p
r
o
ce
s
s
i
n
g
,
tr
ain
in
g
,
a
n
d
test
i
n
g
p
ar
t
o
f
th
e
p
r
o
j
ec
t m
u
ch
m
o
r
e
b
ea
r
ab
le
k
ee
p
in
g
t
i
m
e
co
m
p
lex
i
t
y
in
m
in
d
.
Ov
er
all,
it
w
as
lear
n
ed
t
h
at
b
y
u
s
in
g
li
n
ea
r
r
eg
r
ess
io
n
m
o
d
el
an
d
tak
i
n
g
ea
c
h
b
lan
k
ce
ll
as
a
u
n
iq
u
e
attr
ib
u
te
to
h
an
d
li
n
g
m
is
s
in
g
d
ata,
th
e
d
ata
ac
c
u
r
ac
y
w
as
ab
le
to
b
e
en
h
an
ce
d
b
y
5
%
f
r
o
m
9
3
%
to
9
8
%.
Fu
t
u
r
e
i
m
p
r
o
v
e
m
e
n
ts
w
il
l
in
c
lu
d
e
th
e
u
s
e
o
f
m
o
r
e
co
m
p
lex
m
o
d
els
s
u
c
h
as
u
s
i
n
g
n
e
u
r
al
n
et
w
o
r
k
s
to
s
ee
if
th
e
d
ata
ac
cu
r
ac
y
ca
n
i
n
cr
ea
s
e
to
b
e
ev
en
s
li
g
h
tl
y
b
etter
o
r
if
th
er
e
ar
e
an
y
ch
a
n
g
es
as
w
ell
as
ch
alle
n
g
es
t
h
at
co
m
e
w
i
th
it.
Ot
h
er
f
u
t
u
r
e
i
m
p
r
o
v
e
m
e
n
t
i
n
clu
d
ed
h
o
w
to
i
m
p
le
m
e
n
t
m
is
s
in
g
d
ata
h
an
d
lin
g
tec
h
n
iq
u
es
i
n
o
th
er
m
o
r
e
co
m
p
licated
m
o
d
els
s
u
c
h
as
r
ein
f
o
r
ce
m
en
t
lea
r
n
in
g
u
s
in
g
g
r
ee
d
y
alg
o
r
it
h
m
s
to
test
th
e
ag
en
t
lear
n
in
g
a
n
d
d
if
f
er
e
n
t
m
et
h
o
d
s
in
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
m
e
t
h
o
d
s
s
u
c
h
as
k
-
n
ea
r
es
t
n
ei
g
h
b
o
r
s
(
KNN)
to
s
h
o
w
p
o
ten
tial d
is
cr
ete
an
d
co
n
ti
n
u
o
u
s
attr
ib
u
te
s
[
14
].
5.
CO
NCLU
SI
O
N
Usi
n
g
th
e
co
n
v
e
n
tio
n
al
m
et
h
o
d
s
f
o
r
h
an
d
lin
g
m
i
s
s
i
n
g
d
ata,
th
e
d
ata
w
er
e
r
ed
u
ce
d
to
8
,
6
0
3
en
tr
ies
in
s
tead
o
f
3
0
7
,
5
1
1
en
tr
ies
o
f
th
e
o
r
i
g
in
a
l
d
ataset.
T
h
e
n
o
v
el
al
g
o
r
ith
m
d
i
s
cu
s
s
ed
i
n
t
h
is
p
ap
er
allo
w
s
t
h
e
u
s
a
g
e
o
f
all
th
e
av
ailab
le
d
ata.
T
h
e
n
e
w
m
et
h
o
d
en
h
a
n
ce
d
th
e
ac
cu
r
ac
y
o
f
th
e
m
ac
h
i
n
e
lear
n
in
g
m
o
d
el
test
ed
f
r
o
m
9
3
%
u
s
i
n
g
t
h
e
co
n
v
e
n
ti
o
n
al
alg
o
r
it
h
m
s
to
9
8
%
u
s
in
g
o
u
r
n
o
v
el
al
g
o
r
ith
m
.
T
h
e
n
o
v
el
m
eth
o
d
ex
ten
d
s
f
u
r
t
h
er
t
h
an
j
u
s
t
t
h
i
s
r
esear
ch
as
it
ca
n
b
e
ap
p
lied
in
d
i
f
f
er
en
t
ar
ea
s
i
n
clu
d
i
n
g
tr
ain
in
g
n
et
wo
r
k
s
li
k
e
t
h
e
A
T
V
m
en
tio
n
ed
in
t
h
e
d
is
c
u
s
s
io
n
s
e
ctio
n
ab
o
v
e.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
an
k
s
to
Dr
.
Min
w
o
o
(
J
ac
k
)
L
ee
f
o
r
h
is
g
u
id
a
n
ce
in
t
h
i
s
r
esear
ch
.
RE
F
E
R
E
NC
E
S
[1
]
B.
A
n
g
e
lo
v
,
“
W
o
rk
in
g
w
it
h
M
issin
g
Da
ta
in
M
a
c
h
in
e
L
e
a
rn
in
g
,
”
in
T
o
wa
r
d
s
Da
ta
S
c
ien
c
e
,
2
0
1
7
.
[
O
n
li
n
e
]
.
Av
a
il
a
b
le:
h
tt
p
s://
to
w
a
rd
sd
a
tas
c
ie
n
c
e
.
c
o
m
/
w
o
rk
in
g
-
w
it
h
-
m
is
sin
g
-
d
a
ta
-
in
-
m
a
c
h
in
e
-
lea
rn
in
g
-
9
c
0
a
4
3
0
d
f
4
c
e
.
[2
]
K.
M
a
lad
k
a
r,
“
5
W
a
y
s
to
Ha
n
d
le
M
issi
n
g
V
a
lu
e
s
in
M
a
c
h
in
e
L
e
a
rn
in
g
Da
tas
e
ts,”
2
0
1
8
,
A
v
a
il
a
b
le:
h
tt
p
s:/
/www
.
a
n
a
l
y
ti
c
sin
d
ia
m
a
g
.
c
o
m
/5
-
wa
y
s
-
h
a
n
d
le
-
m
issin
g
-
v
a
lu
e
s
-
m
a
c
h
in
e
-
lea
rn
in
g
-
d
a
tas
e
ts/
.
[3
]
M
.
P
a
m
p
a
k
a
,
G
.
Hu
tch
e
so
n
,
J.
W
il
li
a
m
s,
“
H
a
n
d
li
n
g
m
issin
g
d
a
ta:
a
n
a
ly
sis
o
f
a
c
h
a
ll
e
n
g
in
g
d
a
ta
se
t
u
sin
g
m
u
l
ti
p
le
im
p
u
tatio
n
,
”
I
n
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Res
e
a
rc
h
&
M
e
th
o
d
in
Ed
u
c
a
ti
o
n
,
v
o
l.
3
9
,
n
o
.
1
,
p
p
.
1
9
-
3
7
,
2
0
1
6
.
[4
]
K.
L
.
M
a
sc
o
n
i,
T
.
E.
M
a
tsh
a
,
J.
B.
Ech
o
u
f
f
o
-
T
c
h
e
u
g
u
i,
R.
T
.
Era
s
m
u
s,
A
.
P
.
Ke
n
g
n
e
,
“
Re
p
o
rti
n
g
a
n
d
h
a
n
d
li
n
g
o
f
m
is
sin
g
d
a
ta
in
p
re
d
ictiv
e
re
se
a
rc
h
f
o
r
p
re
v
a
len
t
u
n
d
iag
n
o
se
d
t
y
p
e
2
d
iab
e
tes
m
e
ll
it
u
s:
A
s
y
st
e
m
a
ti
c
re
v
ie
w
,
”
EP
M
A
J
o
u
rn
a
l
,
v
o
l.
6
,
n
o
.
7
,
p
p
.
1
-
1
1
,
2
0
1
5
.
[5
]
A
li
js,
“
Ho
m
e
Cre
d
it
De
f
a
u
lt
Risk
Ka
g
g
le
c
o
m
p
e
ti
ti
o
n
3
r
d
p
lac
e
so
lu
ti
o
n
,
”
Ka
g
g
le
,
2
0
1
8
.
[
O
n
l
in
e
].
A
v
a
il
a
b
le:
h
tt
p
s:/
/www
.
k
a
g
g
l
e
.
c
o
m
/
c
/h
o
m
e
-
c
re
d
it
-
d
e
f
a
u
lt
-
risk
/d
isc
u
ss
io
n
/
6
4
8
3
0
.
[6
]
M
a
x
w
e
ll
,
“
Ho
m
e
Cre
d
it
De
fa
u
lt
Risk
Ka
g
g
le
Co
m
p
e
ti
ti
o
n
2
n
d
p
l
a
c
e
so
lu
ti
o
n
(
tea
m
ik
iri
DS),
”
in
Ka
g
g
le
,
2
0
1
8
.
[
On
li
n
e
]
.
A
v
a
il
a
b
le:
h
tt
p
s://
w
ww
.
k
a
g
g
le.co
m
/c/h
o
m
e
-
c
re
d
it
-
d
e
f
a
u
lt
-
risk
/d
isc
u
ss
io
n
/
6
4
5
1
0
.
[7
]
B.
T
u
n
g
u
z
,
“
Ho
m
e
Cr
e
d
it
De
f
a
u
lt
Risk
Ka
g
g
le
Co
m
p
e
ti
ti
o
n
1
s
t
P
lac
e
S
o
lu
ti
o
n
,
”
in
K
a
g
g
le
,
2
0
1
8
.
[
On
li
n
e
]
.
Av
a
il
a
b
le:
h
tt
p
s://
w
ww
.
k
a
g
g
le.co
m
/c/h
o
m
e
-
c
re
d
it
-
d
e
fa
u
lt
-
risk
/d
isc
u
ss
io
n
/
6
4
8
2
1
.
[8
]
A
.
S
w
a
li
n
,
“
Ho
w
to
Ha
n
d
le M
issin
g
Da
ta,” in
T
o
wa
rd
s D
a
ta
S
c
ien
c
e
,
p
p
.
1
-
1
0
,
2
0
1
8
.
[9
]
Ka
g
g
l
e
“
Ho
m
e
Cr
e
d
it
De
fa
u
lt
Risk
-
Ka
g
g
le,”
2
0
1
8
,
[
On
l
in
e
]
A
v
a
il
a
b
le:
h
tt
p
s://
w
ww
.
k
a
g
g
le.co
m
/c/h
o
m
e
-
c
re
d
it
-
d
e
f
a
u
lt
-
risk
/d
a
ta
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2586
I
A
E
S
I
n
t
J
R
o
b
&
A
u
to
m
,
Vo
l
.
10
,
No
.
2
,
J
u
n
e
2
0
2
1
:
1
2
3
–
132
132
[1
0
]
W
.
Ko
e
h
rse
n
,
“
S
tart
h
e
re
:
A
g
e
n
tl
e
in
tro
d
u
c
ti
o
n
,
”
in
Ka
g
g
le
,
2
0
1
8
.
[
O
n
li
n
e
]
A
v
a
il
a
b
le:
h
tt
p
s:/
/www
.
k
a
g
g
l
e
.
c
o
m
/
w
il
lk
o
e
h
rse
n
/start
-
h
e
re
-
a
-
g
e
n
tl
e
in
tro
d
u
c
ti
o
n
/
n
o
te
b
o
o
k
.
[1
1
]
K.
H.
Eri
a
n
,
S
.
M
h
a
p
a
n
k
a
r,
J.
M
.
Co
n
ra
d
,
a
n
d
S
.
G
a
m
b
il
l,
“
S
y
ste
m
In
teg
ra
ti
o
n
o
v
e
r
a
CA
N
Bu
s
f
o
r
a
S
e
lf
-
Co
n
tr
o
ll
e
d
,
L
o
w
-
Co
st
A
u
to
n
o
m
o
u
s
A
ll
-
terra
in
V
e
h
icle
,
”
2
0
1
9
S
o
u
th
e
a
stCo
n
,
Hu
n
tsv
il
le,
A
L
,
USA
,
p
p
.
1
-
8
,
2
0
1
9
,
d
o
i:
1
0
.
1
1
0
9
/S
o
u
t
h
e
a
stCo
n
4
2
3
1
1
.
2
0
1
9
.
9
0
2
0
3
7
9
.
[1
2
]
K.
H.
Eri
a
n
,
“
S
y
ste
m
In
teg
r
a
ti
o
n
Ov
e
r
a
Ca
n
Bu
s
f
o
r
a
S
e
l
f
-
Co
n
tr
o
ll
e
d
,
L
o
w
-
Co
st
A
u
to
n
o
m
o
u
s
A
ll
-
T
e
rra
in
V
e
h
icle
,
”
M
S
c
th
e
sis
,
T
h
e
Un
iv
e
r
sity
o
f
No
rth
Ca
ro
li
n
a
a
t
Ch
a
rl
o
tt
e
,
p
p
.
1
-
1
8
,
2
0
1
9
.
[1
3
]
K.
H.
E
rian
,
J.
M
.
C
o
n
ra
d
,
“
M
e
a
su
rin
g
d
riv
i
n
g
p
e
rf
o
rm
a
n
c
e
f
o
r
a
n
A
ll
-
T
e
rra
in
V
e
h
icle
in
a
p
a
v
e
d
ro
a
d
in
th
e
w
o
o
d
s,”
2
0
2
0
S
o
u
t
h
e
a
stCo
n
,
Ra
l
e
ig
h
,
NC,
USA
,
p
p
.
1
-
8
,
2
0
2
0
,
d
o
i:
1
0
.
1
1
0
9
/S
o
u
t
h
e
a
stCo
n
4
4
0
0
9
.
2
0
2
0
.
9
2
4
9
7
0
5
.
[1
4
]
G
.
Hu
tch
e
so
n
,
“
M
issi
n
g
d
a
ta:
Da
ta
re
p
lac
e
m
e
n
t
a
n
d
im
p
u
tati
o
n
,
”
J
o
u
rn
a
l
o
f
M
o
d
e
ll
in
g
i
n
M
a
n
a
g
e
me
n
t
,
v
o
l.
7
,
n
o
.
2
,
p
p
.
1
-
1
9
,
2
0
1
2
.
[1
5
]
J.
R.
v
a
n
G
in
k
e
l,
M
.
L
in
ti
n
g
,
R
.
C.
Rip
p
e
,
A
.
v
a
n
d
e
r
Vo
o
r
t,
“
R
e
b
u
tt
i
n
g
e
x
isti
n
g
m
is
c
o
n
c
e
p
ti
o
n
s
a
b
o
u
t
m
u
lt
ip
le
im
p
u
tatio
n
a
s a
m
e
th
o
d
f
o
r
h
a
n
d
l
i
n
g
m
is
sin
g
d
a
ta,”
J
o
u
rn
a
l
o
f
Per
s
o
n
a
li
ty A
ss
e
ss
me
n
t
,
v
o
l.
1
0
2
,
n
o
.
3
,
p
p
.
2
9
7
-
3
0
8
,
2
0
2
0
.
[1
6
]
C
.
M
.
M
u
sil
,
C
.
B.
W
a
rn
e
r,
P
.
K
.
Yo
b
a
s,
a
n
d
S
.
L
.
Jo
n
e
s,
“
A
c
o
m
p
a
riso
n
o
f
i
m
p
u
tatio
n
tec
h
n
i
q
u
e
s
f
o
r
h
a
n
d
li
n
g
m
is
sin
g
d
a
ta,”
W
e
ste
rn
J
o
u
rn
a
l
o
f
Nu
rs
in
g
Res
e
a
rc
h
,
v
o
l
.
2
4
,
n
o
.
7
,
p
p
.
8
1
5
-
8
2
9
,
2
0
0
2
.
[1
7
]
J.
W
.
G
ra
h
a
m
,
P
.
E.
Cu
m
sill
e
,
A
.
E.
S
h
e
v
o
c
k
,
“
M
e
th
o
d
s
f
o
r
h
a
n
d
li
n
g
m
issin
g
d
a
ta,”
Ha
n
d
b
o
o
k
o
f
Psy
c
h
o
lo
g
y
,
S
e
c
o
n
d
Ed
it
io
n
2
,
2
0
1
2
.
[1
8
]
J.
G
a
n
tz
a
n
d
D
.
Re
in
se
l,
“
Ex
trac
ti
n
g
v
a
lu
e
f
ro
m
c
h
a
o
s,”
IDC
IVI
EW
Extra
c
ti
n
g
Va
l
u
e
fro
m
Ch
a
o
s
,
IDC.
1
1
4
2
,
p
p
.
1
-
1
2
,
2
0
1
1
.
[1
9
]
J.
R.
Ch
e
e
m
a
,
“
S
o
m
e
g
e
n
e
ra
l
g
u
id
e
li
n
e
s
f
o
r
c
h
o
o
si
n
g
m
issin
g
d
a
ta
h
a
n
d
li
n
g
m
e
th
o
d
s
in
e
d
u
c
a
ti
o
n
a
l
re
se
a
rc
h
,
”
J
o
u
rn
a
l
o
f
M
o
d
e
rn
Ap
p
li
e
d
S
t
a
ti
s
ti
c
a
l
M
e
th
o
d
s
,
v
o
l
.
1
3
,
n
o
.
2
,
p
p
.
5
3
-
7
5
,
2
0
1
4
.
[2
0
]
S.
M
h
a
p
a
n
k
a
r,
“
A
Na
v
i
g
a
ti
o
n
S
y
ste
m
f
o
r
L
o
w
-
c
o
st
A
u
to
n
o
m
o
u
s
A
ll
-
terra
in
-
v
e
h
icle
s,”
M
.
S
c
.
th
e
sis,
Un
iv
e
rsit
y
o
f
No
rth
Ca
ro
li
n
a
a
t
C
h
a
rlo
t
te,
p
p
.
1
-
7
0
.
2
0
1
9
.
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
K
a
r
i
m
H
.
Er
ia
n
re
c
e
iv
e
d
h
is
B
.
S
.
d
e
g
re
e
in
E
lec
tro
n
ics
a
n
d
C
o
m
m
u
n
ica
ti
o
n
s
En
g
in
e
e
ri
n
g
f
ro
m
Ca
iro
Un
iv
e
rsity
,
E
g
y
p
t,
in
2
0
0
8
a
n
d
h
is
M
.
S
.
d
e
g
re
e
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
f
ro
m
th
e
Un
iv
e
rsit
y
o
f
No
rth
Ca
ro
li
n
a
a
t
Ch
a
rlo
tt
e
in
2
0
1
9
.
He
is
c
u
rre
n
tl
y
p
u
rsu
in
g
h
is
P
h
.
D.
d
e
g
re
e
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
.
F
ro
m
2
0
0
9
to
2
0
1
3
h
e
w
a
s
a
n
e
m
b
e
d
d
e
d
so
f
twa
re
e
n
g
in
e
e
r
a
t
V
a
leo
.
F
ro
m
2
0
1
3
t
o
2
0
1
7
h
e
w
a
s
a
se
n
i
o
r
p
rin
c
i
p
a
l
q
u
o
tati
o
n
e
n
g
in
e
e
r
a
t
V
a
leo
.
F
ro
m
2
0
1
7
to
2
0
1
9
h
e
w
a
s a
F
u
lb
rig
h
t
sc
h
o
lar
a
t
UN
CC.
He
w
o
rk
e
d
a
s
a
tea
c
h
in
g
a
ss
istan
t
f
ro
m
2
0
1
8
to
p
re
se
n
t.
Ped
r
o
H
.
R
e
g
a
la
d
o
r
e
c
e
i
v
e
d
h
is
B.
S
a
n
d
M
.
S
.
d
e
g
re
e
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
f
ro
m
th
e
Un
iv
e
rsit
y
o
f
No
rth
Ca
r
o
li
n
a
a
t
Ch
a
rlo
tt
e
i
n
2
0
1
8
a
n
d
2
0
1
9
,
r
e
sp
e
c
ti
v
e
l
y
.
He
is
c
u
rre
n
tl
y
p
u
rsu
i
n
g
h
is
P
h
.
D.
d
e
g
re
e
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
.
His
re
se
a
rc
h
to
p
ics
in
c
lu
d
e
th
e
In
tern
e
t
o
f
T
h
in
g
s,
we
a
ra
b
le
tec
h
n
o
lo
g
y,
a
n
d
in
telli
g
e
n
t
se
n
so
rs
w
it
h
a
f
o
c
u
s
i
n
m
ix
e
d
re
a
li
t
y
.
F
ro
m
2
0
1
6
to
2
0
2
0
h
e
e
x
p
e
rien
c
e
d
a
v
a
riet
y
o
f
in
tern
sh
ip
r
o
les
f
ro
m
m
u
lt
ip
le
re
s
p
e
c
tab
le
c
o
m
p
a
n
ies
in
c
lu
d
i
n
g
Cisc
o
S
y
ste
m
s,
T
h
e
Bo
e
in
g
Co
m
p
a
n
y
,
M
o
n
ste
r
S
m
a
rt
S
o
lu
t
io
n
s,
Ro
c
k
w
e
ll
A
u
to
m
a
ti
o
n
,
a
n
d
A
lst
o
n
&
Bir
d
a
s
a
n
e
tw
o
rk
e
n
g
in
e
e
r,
e
lec
tri
c
a
l
d
e
sig
n
e
n
g
in
e
e
r,
e
m
b
e
d
d
e
d
sy
ste
m
s
e
n
g
in
e
e
r,
tec
h
n
ica
l
sa
les
a
n
a
ly
ti
c
s
e
n
g
in
e
e
r
,
a
n
d
a
p
a
ten
t
sp
e
c
ialist,
re
sp
e
c
ti
v
e
l
y
.
He
w
o
rk
e
d
a
s
a
tea
c
h
e
r
a
ss
istan
t
f
ro
m
2
0
1
7
to
p
re
se
n
t
a
n
d
is
c
o
n
ti
n
u
in
g
h
is
d
o
c
to
ra
l
d
e
g
re
e
u
n
d
e
r
t
h
e
G
ra
d
u
a
te
A
ss
istan
t
S
u
p
p
o
rt
P
r
o
g
ra
m
a
n
d
NSF
V
e
n
t
u
re
p
rise
I
-
Co
rp
s s
p
o
n
so
re
d
f
u
n
d
i
n
g
.
J
a
m
e
s
M
.
Co
n
r
a
d
re
c
e
i
v
e
d
h
is B
.
S
.
d
e
g
re
e
in
Co
m
p
u
ter
S
c
ien
c
e
f
ro
m
th
e
Un
iv
e
rsity
o
f
Ill
in
o
is,
Urb
a
n
a
,
in
1
9
8
4
a
n
d
h
is
M
.
S
.
a
n
d
P
h
.
D.
d
e
g
re
e
s
in
Co
m
p
u
ter
E
n
g
in
e
e
rin
g
f
ro
m
No
rth
Ca
ro
li
n
a
S
tate
Un
iv
e
rsit
y
in
1
9
8
7
a
n
d
1
9
9
2
,
re
sp
e
c
ti
v
e
l
y
.
F
ro
m
1
9
8
4
t
o
1
9
9
0
h
e
w
a
s
a
so
f
t
wa
re
e
n
g
in
e
e
r
w
it
h
IBM
,
f
ro
m
1
9
9
2
t
o
1
9
9
5
h
e
w
a
s
a
n
a
ss
istan
t
p
ro
f
e
ss
o
r
a
t
th
e
Un
iv
e
rsit
y
o
f
A
r
k
a
n
sa
s,
a
n
d
f
ro
m
1
9
9
7
t
o
2
0
0
3
h
e
w
a
s
a
s
o
f
t
w
a
r
e
e
n
g
in
e
e
r
a
n
d
p
ro
jec
t
m
a
n
a
g
e
r
w
i
th
Eri
c
ss
o
n
a
n
d
S
o
n
y
Eri
c
ss
o
n
.
S
in
c
e
2
0
0
3
h
e
h
a
s
b
e
e
n
a
n
a
ss
o
c
iate
p
ro
f
e
ss
o
r
a
n
d
p
r
o
f
e
s
so
r
w
it
h
th
e
El
e
c
tri
c
a
l
a
n
d
Co
m
p
u
ter
En
g
in
e
e
rin
g
De
p
a
rtm
e
n
t
a
t
t
h
e
Un
iv
e
rsity
o
f
No
rth
Ca
ro
li
n
a
a
t
C
h
a
rlo
t
te.
He
is
th
e
a
u
th
o
r
o
f
e
ig
h
t
b
o
o
k
s
a
n
d
m
o
re
th
a
n
1
4
0
a
rti
c
les
in
t
h
e
a
re
a
s
o
f
e
m
b
e
d
d
e
d
sy
ste
m
s,
ro
b
o
t
ics
,
p
a
ra
ll
e
l
p
ro
c
e
ss
in
g
,
a
n
d
e
n
g
in
e
e
ri
n
g
e
d
u
c
a
ti
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.