I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
A
I
)
V
ol
. 13, No. 4,
D
e
c
e
m
be
r
2024
, pp.
3930
~
3941
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v13.i4.pp
3930
-
3941
3930
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
U
si
n
g
p
at
t
e
r
n
m
i
n
i
n
g
t
o d
e
t
e
r
m
i
n
e
f
i
n
e
c
l
i
m
at
i
c
p
ar
am
e
t
e
r
s f
or
m
ai
z
e
yi
e
l
d
i
n
B
e
n
i
n
S
ou
an
d
P
e
ac
e
G
lo
r
ia
T
ah
i
1
, V
in
as
e
t
an
R
at
h
e
il
H
o
u
n
d
j
i
2
, C
a
s
t
r
o G
b
ê
m
ê
m
al
i
H
ou
n
m
e
n
ou
1
,
R
om
ai
n
G
lè
lè
K
ak
aï
1
1
L
a
bor
a
t
oi
r
e
de
B
i
om
a
t
hé
m
a
t
i
que
s
e
t
d’
E
s
t
i
m
a
t
i
ons
F
or
e
s
t
i
è
r
e
s
, F
a
c
ul
t
y of
A
gr
onom
i
c
S
c
i
e
nc
e
s
,
U
ni
ve
r
s
i
t
y of
A
bom
e
y
-
C
a
l
a
vi
,
C
ot
onou, B
e
ni
n
2
I
ns
t
i
t
ut
de
F
or
m
a
t
i
on e
t
de
R
e
c
he
r
c
he
e
n I
nf
or
m
a
t
i
que
, U
ni
ve
r
s
i
t
y of
A
bom
e
y
-
C
a
l
a
vi
, C
ot
onou, B
e
ni
n
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
D
e
c
3, 2023
R
e
vi
s
e
d
M
a
r
29, 2024
A
c
c
e
pt
e
d
A
pr
17, 2024
This
study
investigates
the
relationships
between
Benin'
s
climate
and
maize
production
to
develop
an
association
rule
algorithm
for
accurate
yield
prediction.
The
datasets
utilized
extend
26
years
(1995
to
2020)
and
i
nclude
climate
and
maize
yield
data
from
five
distri
cts
with
synopt
ic
weather
stations
in
two
agroclimatic
zones
(Sudanian
and
Sudano
-
Guinean).
C
limate
variables
were
combined
with
yield
using
"
year
"
and
"
districts"
to
fi
nd
the
associati
on
rules.
Several
techniqu
es
were
used
to
determin
e
the
corr
elation
between
weather
parameters
and
maize
yields:
support
vector
machines,
K
nearest
neighbor,
artificial
neural
networks,
decision
trees,
and
re
current
neural
networks.
The
most
performed
method
was
the
decisio
n
tree
(R2=
0.998,
mean
squared
error
(
MSE
)
=
0.021,
and
mean
absolute
error
(
MAE
)
=
0.0008).
This
model
is
difficult
to
understand,
though
the
fr
equent
pattern
growth
technique
was
then
applied
to
the
dataset
to
facilitate
the
discovery
of
the
rules.
The
Sudano
-
Guinean
zone
exhibits
high
maize
yields
for
medium
minimum
and
maximum
temperature
values,
r
ainfall,
evapotrans
piration
,
and
humidi
ty.
In
the
Sudanian
zone,
medium
mi
nimum
and
maximum
temperatures
and
maximum
humidity
levels
are
ass
ociated
with
high
maize
yields.
The
discovered
association
rules
showe
d
that
optimizing maize output
might be done
dependa
bly and ef
fectively.
K
e
y
w
o
r
ds
:
A
s
s
oc
ia
ti
on r
ul
e
s
C
li
m
a
ti
c
pa
tt
e
r
n
M
a
c
hi
ne
l
e
a
r
ni
ng
Y
ie
ld
pr
e
di
c
ti
on
Z
e
a
m
a
ys
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
S
oua
nd P
e
a
c
e
G
lo
r
ia
T
a
hi
L
a
bor
a
to
ir
e
de
B
io
m
a
th
é
m
a
ti
que
s
e
t
d’
E
s
ti
m
a
ti
ons
F
or
e
s
ti
è
r
e
s
,
F
a
c
ul
ty
of
A
gr
onomi
c
S
c
ie
nc
e
s
U
ni
ve
r
s
it
y of
A
bom
e
y
-
C
a
la
vi
04 B
P
1525 C
ot
onou, B
é
ni
n
E
m
a
il
:
s
oua
ndt
a
hi
@
gm
a
il
.c
om
1.
I
N
T
R
O
D
U
C
T
I
O
N
A
gr
ic
ul
tu
r
e
is
th
e
pr
im
a
r
y s
our
c
e
of
e
c
onomi
c
gr
ow
th
in
m
os
t
de
ve
lo
pi
ng
c
ount
r
ie
s
.
A
c
hi
e
vi
ng
f
ood
s
e
c
ur
it
y
r
e
qui
r
e
s
,
a
m
ong
ot
he
r
s
,
a
n
in
c
r
e
a
s
e
in
c
r
op
pr
oduc
ti
o
n.
H
ow
e
ve
r
,
s
e
ve
r
a
l
f
a
c
to
r
s
li
m
it
c
r
op
yi
e
ld
s
:
c
li
m
a
te
,
s
oi
l
c
om
pos
it
io
n,
f
e
r
ti
li
z
e
r
us
e
,
r
e
s
our
c
e
a
v
a
il
a
bi
li
ty
,
a
nd
pol
it
ic
a
l
a
nd
s
o
c
io
e
c
onomi
c
v
a
r
ia
bl
e
s
[
1]
.
N
e
ve
r
th
e
le
s
s
,
c
li
m
a
te
c
ha
nge
r
e
m
a
in
s
th
e
pr
im
a
r
y
c
a
u
s
e
r
e
s
tr
ic
ti
ng
a
gr
ic
ul
tu
r
a
l
yi
e
ld
[
2]
.
I
t
le
d
to
s
ig
ni
f
ic
a
nt
di
s
r
upt
io
ns
in
c
e
r
ta
in
r
e
gi
ons
[
3]
,
th
e
r
e
by
a
f
f
e
c
ti
ng
a
gr
ic
ul
tu
r
a
l
pr
oduc
ti
vi
ty
.
N
ow
a
da
ys
,
it
i
s
no
lo
nge
r
pos
s
ib
le
f
or
a
f
a
r
m
e
r
to
pr
oduc
e
w
it
hout
knowing
th
e
yi
e
ld
f
or
e
c
a
s
t,
w
hi
c
h
s
ugg
e
s
ts
a
dopt
in
g
s
m
a
r
t
a
gr
ic
ul
tu
r
e
.
S
ta
ti
s
ti
c
a
l
a
nd
c
om
put
a
ti
ona
l
te
c
hni
que
s
a
r
e
in
c
r
e
a
s
in
gl
y
us
e
d
by
f
a
r
m
e
r
s
,
a
gr
onomi
s
ts
,
a
nd
pol
ic
ym
a
ke
r
s
f
or
de
c
is
io
n
-
m
a
ki
ng
[
4]
.
F
o
r
th
is
r
e
a
s
on,
pr
e
di
c
ti
ng
c
li
m
a
te
c
ha
nge
im
pa
c
ts
on
c
r
op
yi
e
ld
s
,
e
c
onomy, a
nd t
he
e
nvi
r
onm
e
nt
f
r
om
hi
gh
-
pr
e
c
is
io
n m
ode
ls
be
c
om
e
s
ne
c
e
s
s
a
r
y
[
5]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
U
s
in
g patt
e
r
n m
in
in
g t
o de
te
r
m
in
e
f
in
e
c
li
m
at
ic
par
am
e
te
r
s
f
or
m
ai
z
e
y
ie
ld
…
(
Souand P
e
ac
e
G
lo
r
ia
T
ahi
)
3931
C
r
op
yi
e
ld
tr
e
nds
us
ua
ll
y
f
ol
lo
w
th
e
non
-
li
ne
a
r
pr
oc
e
s
s
.
M
a
c
hi
ne
le
a
r
ni
ng
(
M
L
)
m
a
y
th
e
n
be
c
ons
id
e
r
e
d
a
s
a
n
a
ppr
opr
ia
te
m
e
th
od
f
or
m
ode
li
ng
c
r
op
yi
e
ld
s
[
6]
,
[
7]
.
I
nde
e
d,
th
e
pr
og
r
a
m
us
e
s
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
to
m
a
ke
th
e
c
or
r
e
la
ti
ons
be
twe
e
n
th
e
in
put
a
nd
out
put
va
r
ia
bl
e
s
non
-
li
ne
a
r
[
8]
.
N
ow
a
da
ys
,
s
upe
r
vi
s
e
d
ML
te
c
hni
que
s
h
a
ve
ga
in
e
d
s
ig
ni
f
ic
a
nt
r
e
c
og
ni
ti
on
in
c
r
op
yi
e
ld
p
r
e
di
c
ti
on,
le
ve
r
a
gi
ng
s
ophi
s
ti
c
a
te
d
m
ode
ls
s
uc
h
a
s
de
e
p
ne
ur
a
l
n
e
twor
ks
[
9]
–
[
11]
.
T
he
s
e
a
dva
nc
e
d
m
e
th
ods
e
xc
e
l
a
t
c
a
pt
ur
in
g
in
tr
ic
a
te
pa
tt
e
r
ns
w
it
hi
n
a
gr
ic
ul
tu
r
a
l
da
ta
.
N
e
ve
r
th
e
le
s
s
,
th
e
s
e
m
e
th
ods
a
r
e
in
tr
ic
a
te
,
a
nd
th
e
ir
c
om
pl
e
xi
ty
li
e
s
in
th
e
ir
in
te
ns
iv
e
tr
a
in
in
g
r
e
qui
r
e
m
e
nt
s
a
nd
im
pl
e
m
e
nt
a
ti
on,
of
t
e
n
ne
c
e
s
s
it
a
ti
ng
la
r
ge
da
ta
s
e
ts
a
nd s
ubs
ta
nt
ia
l
c
om
put
a
ti
ona
l
r
e
s
our
c
e
s
.
D
e
s
pi
te
th
e
ir
e
f
f
e
c
ti
ve
ne
s
s
,
th
e
ir
la
c
k
of
in
te
r
pr
e
ta
bi
li
ty
c
a
n
pos
e
a
c
ha
ll
e
nge
.
I
nt
e
r
pr
e
ti
ng
r
e
s
ul
ts
r
e
m
a
in
s
a
s
ig
ni
f
ic
a
nt
c
ha
ll
e
nge
w
it
h
ML
te
c
hni
que
s
.
O
ne
s
uc
h
e
xa
m
pl
e
is
de
c
is
io
n
tr
e
e
s
(
D
T
)
,
w
hi
c
h
of
f
e
r
s
om
e
in
te
r
pr
e
ta
bi
li
ty
by
pr
oduc
in
g
s
im
pl
e
r
ul
e
s
in
th
e
f
or
m
of
a
tr
e
e
.
B
ut
,
th
e
m
or
e
c
om
pl
e
x
th
e
tr
e
e
,
th
e
ha
r
de
r
it
is
to
in
te
r
pr
e
t.
H
ow
e
ve
r
,
th
e
r
e
is
unt
a
ppe
d
pot
e
nt
ia
l
f
or
ot
he
r
M
L
te
c
hni
que
s
,
s
uc
h
a
s
a
s
s
oc
ia
ti
on
r
ul
e
s
, w
hi
c
h
a
r
e
unde
r
u
s
e
d
in
c
r
op
yi
e
ld
pr
e
di
c
ti
on.
T
hi
s
p
a
pe
r
f
oc
us
e
s
on
th
e
a
s
s
o
c
ia
ti
on
r
ul
e
s
a
lg
or
it
hm
s
,
us
e
d
to
f
in
d
uns
e
e
n
or
de
s
ir
e
d
p
a
tt
e
r
ns
in
la
r
g
e
da
ta
s
e
t
s
.
W
id
e
ly
u
s
e
d
in
ot
he
r
f
ie
ld
s
s
uc
h
a
s
tr
a
ns
a
c
ti
ona
l
da
ta
s
e
ts
[
12]
,
a
s
s
oc
i
a
ti
on
r
ul
e
s
of
f
e
r
th
e
pos
s
ib
il
it
y
of
di
s
c
ove
r
in
g
c
or
r
e
la
ti
ons
be
tw
e
e
n
va
r
io
us
a
gr
onomi
c
,
m
e
te
or
ol
ogi
c
a
l,
a
nd
e
nvi
r
onm
e
nt
a
l
pa
r
a
m
e
te
r
s
,
to
f
in
d
th
e
r
e
la
ti
ons
a
m
ong
va
r
ia
bl
e
s
.
T
hi
s
s
tu
dy
br
id
ge
s
t
he
ga
p be
twe
e
n t
he
pr
e
s
e
nt
-
da
y s
ta
tu
s
of
t
he
a
r
t
in
c
r
op yie
ld
pr
e
di
c
ti
on a
nd t
he
oppor
tu
ni
ti
e
s
of
f
e
r
e
d
by a
s
s
oc
ia
ti
on r
ul
e
s
.
T
he
r
e
a
r
e
s
e
ve
r
a
l
a
lg
or
it
hm
s
to
s
e
a
r
c
h
a
s
s
oc
ia
ti
on
r
ul
e
s
,
s
u
c
h
a
s
A
pr
io
r
i
(
A
A
)
,
e
c
la
t,
f
r
e
que
nt
pa
tt
e
r
n
gr
ow
th
(
F
P
G
)
,
a
nd
r
a
pi
d
a
s
s
oc
ia
ti
on
r
ul
e
m
in
in
g
(
R
A
R
M
)
.
S
om
e
pr
e
vi
ous
w
or
ks
a
ppl
ie
d
pa
tt
e
r
n
-
f
in
di
ng
te
c
hni
que
s
in
a
gr
ic
ul
tu
r
e
.
F
or
e
xa
m
pl
e
,
K
a
ur
a
nd
A
tt
w
a
l
[
13]
us
e
d
a
s
s
oc
ia
ti
on
r
ul
e
s
te
c
hni
que
s
to
e
va
lu
a
te
th
e
e
f
f
e
c
t
of
te
m
pe
r
a
tu
r
e
a
nd
pr
e
c
ip
it
a
ti
on
on
r
ic
e
y
ie
ld
in
I
ndi
a
.
T
h
e
y
c
onc
lu
d
e
d
th
a
t
r
ic
e
yi
e
ld
de
pe
nds
on
te
m
pe
r
a
tu
r
e
.
I
n
K
e
nya
,
S
il
a
s
a
nd
N
de
r
u
[
14]
us
e
d
th
e
A
pr
io
r
i
a
lg
or
it
hm
to
pr
e
di
c
t
te
a
pr
oduc
ti
on
a
nd
c
onc
lu
de
d
th
a
t
hi
gh
pr
oduc
ti
on
is
li
ke
ly
to
e
xt
e
nd
in
to
J
a
nua
r
y
a
nd
D
e
c
e
m
be
r
.
I
n
P
a
ki
s
t
a
n,
S
upr
o
[
15]
,
us
in
g
a
s
s
o
c
ia
ti
on
r
ul
e
s
a
nd
a
r
ti
f
ic
ia
l
ne
ur
a
l
ne
twor
k
(
A
N
N
)
,
r
e
ve
a
le
d
th
a
t
r
ic
e
yi
e
ld
is
hi
gh
w
h
e
n
th
e
te
m
pe
r
a
tu
r
e
is
hi
gh,
a
nd
hum
id
it
y
i
s
m
e
di
um
.
M
or
e
ov
e
r
,
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
te
m
pe
r
a
tu
r
e
,
r
a
in
f
a
ll
,
s
oi
l
ni
tr
oge
n,
a
nd
r
ic
e
p
r
oduc
ti
on
w
a
s
a
s
c
e
r
ta
in
e
d
by
R
a
o
e
t
al
.
[
1]
us
in
g
A
pr
io
r
i,
E
c
la
t,
a
nd
A
pr
io
r
iT
id
.
T
he
y
f
ound
th
a
t
th
e
A
pr
io
r
i
a
lg
or
it
hm
out
pe
r
f
or
m
e
d
th
e
ot
he
r
two
a
lg
or
it
hm
s
to
f
in
d
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
w
e
a
th
e
r
pa
r
a
m
e
te
r
s
a
nd
m
a
iz
e
yi
e
ld
s
.
S
a
la
nka
r
e
t
al
.
[
16]
us
e
d
A
pr
io
r
i,
F
P
G
,
a
nd
c
lu
s
te
r
in
g
to
p
r
opos
e
a
s
ol
ut
io
n f
or
c
r
op s
ugge
s
ti
on a
t
a
c
r
op s
e
t.
T
he
y ob
s
e
r
ve
d t
ha
t
th
e
pr
opos
e
d a
ppr
oa
c
h e
f
f
e
c
ti
ve
ly
s
ugge
s
te
d t
he
be
s
t
c
r
op t
o gr
ow
ba
s
e
d on e
nvi
r
onm
e
nt
a
l
f
a
c
to
r
s
.
T
hi
s
pa
pe
r
in
ve
s
ti
ga
te
s
th
e
a
s
s
oc
ia
ti
on
r
ul
e
s
be
twe
e
n
c
li
m
a
te
p
a
r
a
m
e
te
r
s
a
nd
yi
e
ld
to
pr
e
di
c
t
m
a
iz
e
yi
e
ld
in
B
e
ni
n
(
W
e
s
t
A
f
r
ic
a
)
.
M
a
i
z
e
is
in
d
e
e
d
a
n
im
por
ta
nt
c
e
r
e
a
l
c
r
op
in
pe
opl
e
’
s
di
e
t
s
w
or
ld
w
id
e
,
pa
r
ti
c
ul
a
r
ly
in
W
e
s
t
A
f
r
ic
a
[
17]
.
A
nnua
l
m
a
iz
e
c
ons
um
pt
io
n
in
B
e
ni
n
is
55
kg/
c
a
pi
ta
[
18]
,
r
a
nki
ng
th
e
c
ount
r
y
a
t
th
e
to
p
of
c
ount
r
ie
s
w
it
h
hi
gh
m
a
iz
e
c
on
s
um
pt
io
n.
H
ow
e
ve
r
,
f
ol
lo
w
in
g
th
e
e
f
f
e
c
ts
of
c
li
m
a
te
c
ha
nge
,
a
gr
a
dua
l
de
c
li
n
e
in
yi
e
ld
s
a
nd
pr
oduc
ti
on
vol
um
e
ha
s
be
e
n
ob
s
e
r
ve
d
in
r
e
c
e
nt
ye
a
r
s
[
18]
.
I
t
is
th
e
r
e
f
or
e
ur
ge
nt
to
de
te
r
m
in
e
th
e
r
ul
e
s
b
e
twe
e
n c
li
m
a
te
pa
r
a
m
e
te
r
s
a
nd
yi
e
ld
.
F
or
th
is
pur
pos
e
,
i)
m
a
iz
e
yi
e
ld
is
pr
e
di
c
te
d
f
r
om
di
f
f
e
r
e
nt
s
upe
r
vi
s
e
d
M
L
m
e
th
ods
,
ii
)
th
e
d
a
ta
s
e
t
is
a
ugm
e
nt
e
d
f
r
om
th
e
b
e
s
t
-
id
e
nt
if
ie
d
m
e
th
od, a
nd i
ii
)
th
e
ge
ne
r
a
te
d da
ta
i
s
us
e
d t
o e
s
ta
bl
i
s
h t
he
a
s
s
oc
ia
ti
on r
ul
e
s
f
r
om
t
he
F
P
G
a
lg
or
it
hm
.
T
he
r
e
s
t
of
th
e
pa
pe
r
is
or
ga
ni
z
e
d
a
s
f
ol
lo
w
s
.
S
e
c
ti
on
2
pr
e
s
e
nt
s
th
e
da
ta
c
ol
le
c
ti
on
m
e
th
odol
ogy,
ge
ne
r
a
ti
on,
a
nd
a
na
ly
s
is
.
T
he
n
s
e
c
ti
on
3
de
a
ls
w
it
h
th
e
m
a
in
r
e
s
ul
ts
obt
a
in
e
d
a
nd
di
s
c
us
s
io
n.
F
in
a
ll
y,
s
e
c
ti
on 4 f
oc
us
e
s
on t
he
c
onc
lu
s
io
n.
2.
M
E
T
H
O
D
2.1. Dat
a c
ol
le
c
t
io
n
an
d
p
r
e
p
r
oc
e
s
s
in
g
2.1.1. Dat
as
e
t
c
on
s
id
e
r
e
d
S
e
c
onda
r
y
da
ta
f
r
om
two
a
gr
oc
li
m
a
ti
c
z
one
s
in
B
e
ni
n
(
S
uda
ni
a
n
a
nd
S
uda
no
-
G
ui
ne
a
n)
a
r
e
us
e
d
to
a
s
s
e
s
s
c
li
m
a
ti
c
f
a
c
to
r
s
a
nd
m
a
iz
e
yi
e
ld
r
e
la
ti
ons
hi
ps
in
B
e
ni
n.
T
he
s
e
da
t
a
w
e
r
e
obt
a
in
e
d
ove
r
26
ye
a
r
s
f
r
om
1995
to
2020
in
f
iv
e
s
ynopti
c
w
e
a
th
e
r
s
t
a
ti
ons
lo
c
a
t
e
d
in
th
e
S
uda
no
-
G
ui
ne
a
n
z
one
(
B
ohi
c
on
a
nd
S
a
ve
di
s
tr
ic
ts
)
a
nd
S
uda
ni
a
n
z
one
(
P
a
r
a
kou,
N
a
ti
ti
ngou,
a
nd
K
a
ndi
di
s
tr
ic
ts
)
.
D
a
ta
c
ove
r
va
r
ia
bl
e
s
s
uc
h
a
s
e
va
pot
r
a
ns
pi
r
a
ti
on
(
m
m
)
,
s
uns
tr
oke
(
h)
,
r
a
in
f
a
ll
(
m
m
)
,
m
in
im
u
m
a
nd
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
s
(
°C
)
,
m
in
im
um
a
nd
m
a
xi
m
um
hum
id
it
y
(
%
)
,
a
nd
ove
r
a
ll
m
a
iz
e
yi
e
ld
(
kg/
h
a
)
.
T
he
c
li
m
a
te
da
ta
(
te
m
pe
r
a
tu
r
e
,
hum
id
it
y,
r
a
in
f
a
ll
,
s
uns
tr
oke
,
e
va
pot
r
a
ns
pi
r
a
ti
on)
a
r
e
obt
a
in
e
d
a
t
th
e
“
A
ge
nc
e
pour
la
S
e
c
ur
it
é
de
la
N
a
vi
ga
ti
on
A
é
r
ie
nne
e
n A
f
r
iq
ue
(
A
S
E
C
N
A
)
”
i
n
B
e
ni
n. T
he
a
nnua
l
m
a
iz
e
yi
e
ld
da
ta
l
in
ke
d t
o t
he
di
s
tr
ic
ts
a
r
e
obt
a
in
e
d a
t
th
e
“
D
ir
e
c
ti
on
de
s
S
ta
ti
s
ti
que
s
A
gr
ic
ol
e
s
(
D
S
A
)
”
.
F
ur
th
e
r
m
or
e
,
da
il
y
da
t
a
on
c
li
m
a
te
va
r
ia
bl
e
s
a
nd
a
nnua
l
yi
e
ld
s
a
r
e
a
ls
o
obt
a
in
e
d
f
r
om
D
S
A
.
I
n
B
e
ni
n,
m
a
iz
e
is
c
ul
ti
va
t
e
d
f
r
om
m
id
-
A
pr
i
l
to
m
id
-
O
c
to
be
r
(
7
m
ont
hs
)
.
H
e
nc
e
,
onl
y
da
ta
c
ove
r
in
g
th
e
pr
oduc
ti
on
pe
r
io
d
a
r
e
c
ol
le
c
te
d
,
1
,
503
da
il
y
c
li
m
a
te
r
e
c
or
ds
.
F
r
om
th
e
da
il
y
da
ta
,
a
ve
r
a
ge
d
m
ont
hl
y
c
li
m
a
te
da
ta
is
a
ggr
e
g
a
te
d
to
yi
e
ld
da
ta
f
or
e
a
c
h
ye
a
r
a
nd
th
e
di
s
tr
ic
t.
F
in
a
ll
y,
875
obs
e
r
va
ti
ons
a
r
e
ut
il
iz
e
d
f
or
th
e
a
na
ly
s
is
.
F
ig
ur
e
1
di
s
pl
a
y
s
t
he
di
s
tr
ib
ut
io
n
of
th
e
c
li
m
a
te
pa
r
a
m
e
te
r
s
by
di
s
tr
ic
ts
.
E
va
pot
r
a
ns
pi
r
a
ti
on
w
a
s
hi
gh
in
th
e
K
a
ndi
di
s
tr
ic
t
s
ho
w
n
in
F
ig
ur
e
1
(
a
)
,
w
hi
le
r
a
in
f
a
ll
w
a
s
hi
ghe
r
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
. 13, No. 4, D
e
c
e
m
be
r
2024:
3930
-
3941
3932
th
e
N
a
ti
ti
ngou
a
nd
K
a
ndi
di
s
tr
ic
ts
s
how
n
in
F
ig
ur
e
1(
b)
.
S
uns
tr
oke
s
how
n
in
F
ig
ur
e
1(
c
)
a
nd
te
m
pe
r
a
tu
r
e
s
how
n
in
F
ig
ur
e
s
1(
d)
a
nd
1(
e
)
ove
r
th
e
la
s
t
26
ye
a
r
s
w
e
r
e
a
ls
o
hi
gh
in
th
e
K
a
ndi
di
s
tr
ic
t.
B
ohi
c
on
a
nd
S
a
vè
ha
d
th
e
hi
ghe
s
t
m
a
xi
m
um
hum
id
it
y
s
how
n
in
F
ig
ur
e
1(
f
)
,
w
hi
le
N
a
ti
ti
ngou,
S
a
vè
,
a
nd
B
ohi
c
on
r
e
c
or
de
d
th
e
hi
ghe
s
t
m
in
im
um
hum
id
it
y
le
ve
ls
s
how
n
in
F
ig
ur
e
1(
g)
.
R
e
ga
r
di
ng
m
a
iz
e
yi
e
ld
,
K
a
ndi
ha
d
th
e
hi
ghe
s
t
va
lu
e
,
f
ol
lo
w
e
d by B
ohi
c
on a
nd N
a
ti
ti
ngou
s
how
n i
n
F
ig
ur
e
1(
h)
.
(
a
)
(
b)
(
d)
(
c
)
(
e
)
(f)
(
g)
(
h)
F
ig
ur
e
1. T
he
di
s
tr
ib
ut
io
n of
c
li
m
a
te
a
nd yie
ld
pa
r
a
m
e
te
r
s
f
or
t
he
1995
-
2020 pe
r
io
d l
in
ke
d t
o e
a
c
h di
s
tr
ic
t.
(
a
)
e
va
pot
r
a
ns
pi
r
a
ti
on, (
b)
r
a
in
f
a
ll
,
(
c
)
s
uns
tr
oke
, (
d)
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
, € mi
ni
m
um
t
e
m
pe
r
a
tu
r
e
,
(
f
)
m
a
xi
m
um
humi
d
it
y, (
g)
m
in
im
um
humi
di
ty
,
a
nd
(
h
)
M
a
iz
e
yi
e
ld
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
U
s
in
g patt
e
r
n m
in
in
g t
o de
te
r
m
in
e
f
in
e
c
li
m
at
ic
par
am
e
te
r
s
f
or
m
ai
z
e
y
ie
ld
…
(
Souand P
e
ac
e
G
lo
r
ia
T
ahi
)
3933
a
)
M
a
iz
e
yi
e
ld
da
ta
s
e
t
T
he
m
a
iz
e
yi
e
ld
tr
e
nd
by
he
c
ta
r
e
f
or
26
ye
a
r
s
is
s
im
il
a
r
f
o
r
B
o
hi
c
on,
K
a
ndi
,
a
nd
N
a
ti
ti
ngou.
T
hi
s
is
a
dow
nw
a
r
d t
r
e
nd f
or
K
a
ndi
, w
it
h t
he
pe
a
k note
d i
n 1997
(
2
,
00
0 kg/ha
)
, w
hi
le
B
ohi
c
on
pr
e
s
e
nt
s
a
c
ont
r
a
s
ti
ng
tr
e
nd
w
it
h
3
pe
a
ks
in
1997
(
1
,
480
kg/
ha
)
,
2007
(
1
,
630
kg/
h
a
)
,
a
nd
2015
(
1
,
900
kg/
ha
)
ha
)
,
r
e
s
pe
c
ti
ve
ly
s
how
n
in
F
ig
ur
e
2.
M
a
iz
e
yi
e
ld
in
N
a
ti
ti
ngou
a
ls
o
s
how
s
a
bi
m
oda
l
tr
e
nd,
w
it
h
pe
a
ks
in
2007
(
1
,
700
kg/
ha
)
a
nd 2012 (
1
,
800 kg/ha
)
. P
a
r
a
kou a
nd
s
a
ve
r
e
c
or
de
d t
he
l
ow
e
s
t
m
a
iz
e
yi
e
ld
s
how
n i
n
F
ig
ur
e
2.
F
ig
ur
e
2. T
e
m
por
a
l
va
r
ia
ti
on i
n m
a
iz
e
yi
e
ld
i
n K
a
ndi
, P
a
r
a
kou, S
a
ve
, N
a
ti
ti
ngou, a
nd B
ohi
c
on
b)
T
r
e
nd of
c
li
m
a
ti
c
pa
r
a
m
e
te
r
s
i
n t
he
pr
oduc
ti
on s
e
a
s
on f
r
om
1995 to 2020 i
n t
he
s
tu
dy a
r
e
a
s
D
ur
in
g
th
e
la
s
t
26
ye
a
r
s
,
K
a
ndi
h
a
s
r
e
gi
s
te
r
e
d
a
hi
gh
pot
e
nt
ia
l
f
or
e
va
pot
r
a
ns
pi
r
a
ti
on
c
om
pa
r
e
d
to
th
e
ot
he
r
di
s
tr
ic
ts
s
how
n
in
F
ig
ur
e
3(
a
)
.
A
m
a
r
ke
d
in
c
r
e
a
s
e
w
a
s
r
e
c
or
de
d
in
1995
a
nd
2000,
w
he
r
e
a
s
a
pr
onounc
e
d
de
c
li
ne
w
a
s
r
e
gi
s
te
r
e
d
in
2001.
I
t
w
a
s
c
ons
ta
nt
f
r
om
1995
to
2011
(
E
T
P
=
2.286
m
m
)
be
f
or
e
r
e
a
c
hi
ng
a
pe
a
k
of
5.1
m
m
in
2020.
O
v
e
r
a
ll
,
K
a
ndi
s
how
e
d
t
he
hi
ghe
s
t
e
va
pot
r
a
ns
pi
r
a
ti
on
va
lu
e
s
,
w
h
e
r
e
a
s
B
ohi
c
on
r
e
c
or
de
d
th
e
lo
w
e
s
t
tr
e
nds
.
R
e
g
a
r
di
ng
th
e
r
a
in
f
a
ll
,
s
im
il
a
r
tr
e
nds
w
e
r
e
not
ic
e
d
f
or
th
e
di
s
tr
ic
ts
c
ons
id
e
r
e
d.
P
a
r
a
kou
a
nd
N
a
ti
ti
ngou
r
e
c
or
de
d
r
e
la
ti
ve
ly
hi
gh
r
a
in
f
a
ll
va
lu
e
s
,
w
he
r
e
a
s
s
a
ve
a
nd
B
ohi
c
on s
how
th
e
lo
w
e
s
t
tr
e
nd
s
s
ho
w
n
in
F
ig
ur
e
3(
b)
.
S
uns
tr
oke
s
how
s
th
e
s
a
m
e
tr
e
nds
a
s
e
v
a
pot
r
a
ns
pi
r
a
ti
on
s
how
n
in
F
ig
ur
e
3(
c
)
.
F
r
om
1995
to
2000,
S
a
ve
,
B
ohi
c
on,
K
a
ndi
,
a
nd
P
a
r
a
kou
r
e
c
or
de
d
th
e
lo
w
e
s
t
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
va
lu
e
s
`
unde
r
25
°C
,
a
nd
s
im
il
a
r
a
nd
f
lu
c
tu
a
ti
ng
t
r
e
nds
be
twe
e
n
32
°C
a
nd
35
°C
f
r
om
2000
t
o
2020
s
how
n
in
F
ig
ur
e
3(
d)
.
R
e
ga
r
di
ng
th
e
m
in
im
um
te
m
pe
r
a
tu
r
e
s
,
S
a
ve
,
N
a
ti
ti
ngou,
a
nd
P
a
r
a
kou
s
how
s
im
il
a
r
t
r
e
nds
ove
r
t
he
a
gr
ic
ul
tu
r
a
l
s
e
a
s
ons
f
or
t
he
l
a
s
t
26 ye
a
r
s
, w
he
r
e
a
s
B
ohi
c
on a
nd K
a
ndi
r
e
c
or
de
d s
im
il
a
r
f
lu
c
tu
a
ti
ons
s
how
n
in
F
ig
u
r
e
3(
e
)
.
H
ow
e
ve
r
,
th
e
m
in
i
m
um
te
m
pe
r
a
tu
r
e
s
r
e
m
a
in
e
d
hi
ghe
r
in
K
a
ndi
c
om
pa
r
e
d
to
ot
he
r
di
s
tr
ic
ts
.
T
he
tr
e
nd
w
a
s
r
e
la
ti
ve
ly
hi
gh
f
or
th
e
B
ohi
c
on
di
s
tr
ic
t
r
e
ga
r
di
ng
th
e
m
a
xi
m
um
a
nd
m
in
im
um
hum
id
it
y.
D
ur
in
g
2013,
m
a
xi
m
um
hum
id
it
y
in
S
a
vè
pe
a
ke
d
a
t
96.64%
s
how
n
in
F
ig
ur
e
3(
f
)
,
w
hi
le
m
in
im
um
hum
id
it
y
pe
a
ke
d
in
2011
a
nd
2014
s
how
n
in
F
ig
ur
e
3(
g)
.
S
im
il
a
r
ly
,
th
e
m
a
xi
m
um
a
nd
m
in
im
u
m
hum
id
it
y
w
e
r
e
r
e
la
ti
ve
ly
lo
w
in
th
e
K
a
ndi
di
s
tr
ic
t.
I
n
N
a
ti
ti
ngo
u,
m
a
xi
m
um
hum
id
i
ty
ha
s
be
e
n
pr
ogr
e
s
s
iv
e
ly
de
c
r
e
a
s
in
g s
in
c
e
2008.
2.1.2. Dat
a ge
n
e
r
at
e
d
N
e
w
w
e
a
th
e
r
a
nd
yi
e
ld
d
a
ta
w
e
r
e
ge
ne
r
a
te
d
f
or
e
a
c
h
of
th
e
f
i
ve
di
s
tr
ic
ts
in
a
gr
o
-
c
li
m
a
ti
c
z
one
s
to
pr
ovi
de
r
e
li
a
bl
e
a
s
s
oc
ia
ti
on
r
ul
e
s
f
or
hi
gh
m
a
iz
e
yi
e
ld
.
T
hi
s
s
im
ul
a
ti
on
us
e
d
th
e
s
upe
r
vi
s
e
d
le
a
r
ni
ng
m
ode
l
w
it
h
R
2
=0
.
99
a
nd
M
A
E
=0
.
0008
.
T
hus
,
th
e
in
put
v
a
r
ia
bl
e
s
o
f
ge
ne
r
a
te
d
da
ta
(
c
li
m
a
te
pa
r
a
m
e
te
r
s
)
f
ol
lo
w
di
s
tr
ib
ut
io
ns
a
nd pa
r
a
m
e
te
r
s
s
im
il
a
r
t
o t
hos
e
i
n t
he
a
c
tu
a
l
da
ta
.
T
e
n t
hous
a
nd (
10,000)
obs
e
r
va
ti
ons
ha
ve
b
e
e
n
ge
ne
r
a
te
d
f
or
e
a
c
h
di
s
tr
ic
t
in
bot
h
a
gr
o
-
c
li
m
a
ti
c
z
one
s
.
D
a
ta
va
r
ia
bl
e
s
w
e
r
e
e
nc
ode
d
in
or
di
na
l
(
lo
w
,
m
e
di
um
,
a
nd
hi
gh)
ba
s
e
d
on
th
e
f
ix
e
d
th
r
e
s
hol
d
va
lu
e
s
.
O
bs
e
r
va
ti
on
s
w
it
h
hi
gh
yi
e
ld
w
e
r
e
s
e
le
c
te
d
f
r
om
e
a
c
h
ge
ne
r
a
te
d
da
ta
ba
s
e
to
bui
ld
th
e
r
ul
e
s
.
T
he
th
r
e
s
hol
d
va
lu
e
s
f
or
a
ll
va
r
ia
bl
e
s
a
r
e
gi
ve
n
by
th
e
a
gr
oe
c
ol
ogi
c
a
l
z
one
i
n T
a
bl
e
1.
2.1.3. Dat
a p
r
e
p
r
oc
e
s
s
in
g
T
he
r
ul
e
s
in
th
e
two
a
gr
o
-
c
li
m
a
ti
c
z
one
s
w
e
r
e
obt
a
in
e
d
by
a
ppl
yi
ng
s
upe
r
vi
s
e
d
a
nd
un
s
upe
r
vi
s
e
d
ML
m
ode
ls
.
T
he
s
up
e
r
vi
s
e
d
m
ode
ls
opt
im
iz
e
m
a
iz
e
yi
e
ld
,
f
in
d
s
ui
ta
bl
e
r
e
la
ti
ons
hi
ps
be
twe
e
n
c
li
m
a
te
pa
r
a
m
e
te
r
s
a
nd
yi
e
ld
,
a
nd
id
e
nt
if
y
th
e
be
s
t
m
ode
l.
T
he
be
s
t
m
ode
l
id
e
nt
if
ie
d
w
a
s
not
e
a
s
il
y
in
te
r
pr
e
ta
bl
e
.
T
hus
,
th
is
m
ode
l
w
it
h
r
a
ndom
pa
r
ts
w
a
s
us
e
d
to
c
om
pl
e
te
th
e
a
va
il
a
bl
e
da
ta
.
T
he
r
e
f
or
e
,
th
e
uns
upe
r
vi
s
e
d
m
ode
l
w
a
s
us
e
d t
o d
e
te
r
m
in
e
t
he
r
ul
e
s
t
ha
t
a
r
e
e
a
s
il
y i
nt
e
r
pr
e
ta
bl
e
a
nd unde
r
s
ta
nda
bl
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
. 13, No. 4, D
e
c
e
m
be
r
2024:
3930
-
3941
3934
B
e
f
or
e
pe
r
f
or
m
in
g
th
e
s
upe
r
vi
s
e
d
m
ode
li
ng,
a
c
or
r
e
la
ti
on
a
na
ly
s
is
be
twe
e
n
in
de
pe
nde
nt
a
nd
de
pe
nde
nt
va
r
ia
bl
e
s
w
a
s
pe
r
f
or
m
e
d.
N
o
c
li
m
a
ti
c
va
r
ia
bl
e
s
w
e
r
e
c
or
r
e
la
te
d
to
m
a
iz
e
yi
e
ld
.
A
n
a
ddi
ti
ona
l
c
or
r
e
la
ti
on
a
na
ly
s
is
w
a
s
pe
r
f
or
m
e
d
a
m
ong
th
e
in
de
pe
nde
nt
va
r
ia
bl
e
s
.
T
he
th
r
e
s
hol
d
of
c
or
r
e
la
ti
on
w
a
s
f
ix
e
d
to
80%
.
U
s
in
g
th
is
th
r
e
s
hod,
th
e
m
in
im
um
hum
id
it
y
w
a
s
e
li
m
in
a
te
d
f
r
om
th
e
pr
e
di
c
to
r
s
.
M
is
s
in
g
da
il
y
da
ta
w
a
s
s
ubj
e
c
te
d
to
s
im
pl
e
im
put
a
ti
on
be
f
or
e
a
ggr
e
ga
ti
on,
a
nd
th
e
out
li
e
r
s
w
e
r
e
s
ubs
e
que
nt
ly
r
e
m
ove
d
f
ol
lo
w
in
g
th
e
in
te
r
qua
r
ti
le
r
a
nge
a
ppr
oa
c
h.
D
a
t
a
w
e
r
e
th
e
n
s
ta
nda
r
di
z
e
d
a
nd
pa
r
ti
ti
one
d
us
in
g
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on. T
he
m
ode
l
s
w
e
r
e
t
r
a
in
e
d on 700 obs
e
r
va
ti
ons
a
nd va
li
da
te
d on the
r
e
m
a
in
in
g 175 da
ta
.
(
a
)
(
b)
(
c
)
(
d)
(
e
)
(f)
(
g)
F
ig
ur
e
3. V
a
r
ia
ti
ons
i
n w
e
a
th
e
r
pa
r
a
m
e
te
r
s
ove
r
t
im
e
i
n K
a
ndi
,
P
a
r
a
kou, S
a
ve
, N
a
ti
ti
ngou, a
nd B
ohi
c
on.
(
a
)
e
va
pot
r
a
ns
pi
r
a
ti
on, (
b)
r
a
in
f
a
ll
,
(
c
)
s
uns
tr
oke
, (
d)
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
, (
e
)
m
in
im
um
t
e
m
pe
r
a
tu
r
e
,
(
f
)
m
a
xi
m
um
humi
d
it
y,
a
nd
(
g)
m
in
im
u
m
humi
di
ty
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
U
s
in
g patt
e
r
n m
in
in
g t
o de
te
r
m
in
e
f
in
e
c
li
m
at
ic
par
am
e
te
r
s
f
or
m
ai
z
e
y
ie
ld
…
(
Souand P
e
ac
e
G
lo
r
ia
T
ahi
)
3935
T
a
bl
e
1.
T
hr
e
s
hol
d va
lu
e
s
f
or
di
s
c
r
e
ti
z
a
ti
on
A
t
t
r
i
but
e
s
L
ow
M
e
di
um
H
i
gh
S
uda
ni
a
n z
one
ETP
<
1
.
618
1.618
-
4.801
≥4
.
801
RR
<
2
.
163
2.163
-
8.206
≥8
.
206
S
un
<
5
.
221
5.221
-
7.883
≥7
.
883
T
m
a
x
<
29
.
371
29.371
-
35.971
≥
35
.
971
T
m
i
n
<
21
.
337
21.337
-
24.015
≥
24
.
015
U
m
a
x
<
87
.
324
87.324
-
102.28
≥
102
.
28
Y
i
e
l
d
<
1076
.
614
1076.614
-
1459.829
≥
1459
.
829
S
uda
no
-
G
ui
ne
a
n z
one
ETP
<
2
.
513
2.513
-
3.628
≥3
.
628
RR
<
2
.
314
2.314
-
6.767
≥6
.
767
S
un
<
3
.
891
3.891
-
6.952
≥6
.
952
T
m
a
x
<
29
.
836
29.836
-
33.599
≥
33
.
599
T
m
i
n
<
22
.
088
22.088
-
23.88
≥
23
.
88
U
m
a
x
<
93
.
497
93.497
-
96.302
≥
96
.
302
Y
i
e
l
d
<
1016
.
311
1016.311
-
1279.765
≥
1279
.
765
2.2. P
r
e
d
ic
t
io
n
of
m
ai
z
e
yi
e
ld
s
w
it
h
s
u
p
e
r
vi
s
e
d
m
ac
h
in
e
l
e
a
r
n
in
g t
e
c
h
n
iq
u
e
s
ML
a
lg
or
it
hm
s
a
r
e
w
id
e
ly
r
e
now
ne
d
f
or
th
e
ir
a
bi
li
ty
to
m
ode
l
non
-
li
ne
a
r
c
or
r
e
la
ti
ons
be
twe
e
n
in
put
a
nd
out
put
va
r
ia
bl
e
s
.
T
o
unde
r
s
t
a
nd
th
e
e
xi
s
ti
ng
non
-
li
ne
a
r
r
e
la
ti
ons
hi
p
be
twe
e
n
c
li
m
a
te
pa
r
a
m
e
te
r
s
a
nd
m
a
iz
e
yi
e
ld
,
f
iv
e
(
5)
s
upe
r
vi
s
e
d
ML
te
c
hni
que
s
w
e
r
e
e
xpl
or
e
d
on
th
e
da
ta
s
e
t
in
P
yt
hon
3.8.11
s
of
twa
r
e
t
o
f
in
d
r
e
la
ti
ons
hi
ps
be
twe
e
n
a
ggr
e
ga
te
d
yi
e
ld
a
nd
w
e
a
th
e
r
f
e
a
tu
r
e
s
.
T
he
s
e
m
e
th
ods
w
e
r
e
k
-
ne
a
r
e
s
t
ne
ig
hbor
(
K
N
N
)
,
s
uppor
t
ve
c
to
r
m
a
c
hi
ne
(
S
V
M
)
,
D
T
,
A
N
N
w
it
h
m
ul
ti
la
ye
r
pe
r
c
e
pt
r
on,
a
nd
r
e
c
ur
r
e
nt
ne
ur
a
l
ne
twor
k
(
R
N
N
)
w
it
h
lo
ng
s
hor
t
-
te
r
m
m
e
m
or
y.
I
nput
va
r
ia
bl
e
s
w
e
r
e
s
ta
nda
r
di
z
e
d
be
f
or
e
m
ode
li
ng.
T
he
da
t
a
w
a
s
di
vi
de
d
in
to
tr
a
in
in
g
a
nd
te
s
ti
ng
u
s
in
g
f
iv
e
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on.
T
he
v
a
r
io
us
m
e
th
ods
w
e
r
e
a
s
s
e
s
s
e
d
us
in
g
th
e
te
s
t
d
a
ta
a
f
te
r
be
in
g
tr
a
in
e
d
on
th
e
tr
a
in
in
g
s
e
t.
T
h
e
“
S
c
ik
i
t
-
L
e
a
r
n”
li
br
a
r
y
w
a
s
us
e
d
f
or
th
e
K
N
N
,
S
V
M
r
e
gr
e
s
s
io
n, a
nd D
T
m
ode
ls
, w
hi
le
“
K
e
r
a
s
”
w
a
s
u
s
e
d f
or
t
he
A
N
N
a
nd
R
N
N
m
ode
ls
.
T
he
f
it
ti
ng
of
th
e
m
ode
ls
to
a
ggr
e
ga
te
out
put
s
w
a
s
s
im
pl
e
.
T
hi
s
is
be
c
a
us
e
e
a
c
h
out
put
c
or
r
e
s
pond
s
to
a
s
e
t
of
in
put
ve
c
to
r
s
.
T
he
pr
obl
e
m
w
a
s
r
e
s
ol
ve
d
by
ge
ne
r
a
ti
ng
a
n
a
r
ti
f
ic
ia
l
out
put
f
or
e
a
c
h
in
put
ve
c
to
r
.
T
he
s
e
a
r
ti
f
ic
ia
l
out
put
s
w
e
r
e
d
e
f
in
e
d
a
s
th
e
m
e
a
n
of
th
e
out
put
va
lu
e
th
r
oughout
th
e
a
ggr
e
ga
te
c
ol
le
c
ti
ons
to
w
hi
c
h t
he
y be
lo
ng. T
he
f
ol
lo
w
in
g s
ub
s
e
c
ti
ons
pr
ovi
de
a
n ov
e
r
vi
e
w
of
t
he
m
ode
ls
us
e
d f
or
t
he
pr
e
di
c
ti
on.
2.2.1. Over
vi
e
w
of
t
h
e
m
od
e
ls
u
s
e
d
A
s
m
e
nt
io
ne
d
a
bove
,
f
iv
e
a
lg
or
it
hm
s
w
e
r
e
c
ons
id
e
r
e
d:
K
N
N
,
S
V
M
,
D
T
,
A
N
N
,
a
nd
R
N
N
.
T
he
s
e
a
lg
or
it
hm
s
a
r
e
a
non
-
pa
r
a
m
e
tr
ic
ML
m
e
th
od
us
e
d t
o unde
r
s
ta
n
d t
he
non
-
li
ne
a
r
l
in
k be
twe
e
n t
he
va
r
ia
bl
e
s
t
ha
t
a
r
e
in
put
a
nd
out
put
.
I
n
K
N
N
r
e
gr
e
s
s
io
n,
pr
e
di
c
ti
on
is
th
e
m
e
a
n
of
th
e
k
ne
a
r
e
s
t
ne
ig
hbor
va
lu
e
s
[
19]
.
S
V
M
goa
l
is
to
m
in
im
iz
e
th
e
p
r
e
di
c
ti
on
bi
a
s
by
f
in
di
ng
a
n
e
xc
e
ll
e
nt
hype
r
-
pl
a
ne
to
r
e
duc
e
di
f
f
e
r
e
nc
e
s
be
twe
e
n
pr
e
di
c
te
d
a
nd
ob
s
e
r
ve
d
va
lu
e
s
[
20]
.
A
D
T
i
s
a
s
upe
r
vi
s
e
d
non
-
pa
r
a
m
e
tr
ic
le
a
r
ni
ng
a
ppr
oa
c
h
th
a
t
f
or
e
c
a
s
t
s
th
e
va
lu
e
of
a
ta
r
ge
t
va
r
ia
bl
e
by
tr
a
in
in
g
on
s
im
pl
e
de
c
i
s
io
n
r
ul
e
s
ba
s
e
d
on
da
ta
c
ha
r
a
c
te
r
is
ti
c
s
.
T
h
e
A
N
N
i
s
ba
s
e
d
on
c
onc
e
pt
ua
l
ne
ur
ons
[
21]
,
[
22]
.
I
t
c
ons
is
ts
of
th
e
in
put
,
hi
dde
n,
a
nd
out
put
la
ye
r
s
of
ne
ur
ons
.
I
n
th
is
w
or
k,
A
N
N
w
it
h
m
ul
ti
la
ye
r
pe
r
c
e
pt
r
on
a
r
c
hi
te
c
tu
r
e
i
s
us
e
d.
R
N
N
is
a
ls
o
u
s
e
d;
it
is
a
n
ANN
c
la
s
s
w
h
e
r
e
c
onne
c
ti
ons
be
twe
e
n
th
e
node
s
f
or
m
a
gr
a
ph
a
lo
ng
a
ti
m
e
s
e
que
nc
e
[
23]
.
R
N
N
c
a
n
pr
oc
e
s
s
a
lo
ng
s
e
que
nc
e
of
in
put
va
r
ia
bl
e
s
us
in
g
th
e
ir
m
e
m
or
y.
I
ts
a
r
c
hi
te
c
tu
r
e
c
ons
is
ts
of
in
put
la
ye
r
s
,
hi
dde
n
la
ye
r
s
,
a
nd
out
pu
t
la
ye
r
s
. E
a
c
h l
a
ye
r
i
s
i
nde
pe
nd
e
nt
of
t
he
ot
he
r
. N
o l
a
ye
r
s
to
r
e
s
t
he
pr
e
c
e
di
ng output
s
.
2.2.2. Com
p
ar
is
on
c
r
it
e
r
ia
T
o
e
va
lu
a
te
th
e
di
f
f
e
r
e
nt
s
upe
r
vi
s
e
d
m
ode
ls
us
e
d,
th
e
c
r
os
s
-
va
li
da
ti
on
(
C
V
)
te
c
hni
que
w
a
s
us
e
d.
D
ue
to
it
s
e
a
s
e
of
us
e
,
a
ppl
ic
a
bi
li
ty
,
a
nd
e
f
f
e
c
ti
ve
ne
s
s
in
pr
e
ve
nt
in
g
th
e
ove
r
-
f
it
ti
ng
pr
obl
e
m
,
it
is
a
n
of
te
n
e
m
pl
oye
d
a
lg
or
it
hm
s
e
le
c
ti
on
te
c
hni
que
[
24]
,
[
25]
.
T
h
e
be
s
t
m
ode
l
is
w
id
e
ly
a
gr
e
e
d
to
b
e
th
e
one
w
it
h
th
e
m
in
im
um
e
s
ti
m
a
ti
on
e
r
r
or
.
U
s
in
g
th
e
c
oe
f
f
ic
ie
nt
of
de
te
r
m
in
a
ti
on
(
R
2)
,
m
e
a
n
s
qua
r
e
e
r
r
or
(
M
S
E
)
,
m
e
a
n
a
bs
ol
ut
e
e
r
r
or
(
M
A
E
)
,
a
nd
r
oot
m
e
a
n
s
qua
r
e
e
r
r
or
(
R
M
S
E
)
,
th
e
m
ode
ls
'
pe
r
f
or
m
a
nc
e
w
a
s
e
va
lu
a
te
d.
R
2
in
di
c
a
te
s
th
e
li
ne
a
r
r
e
la
ti
ons
hi
p
be
twe
e
n
th
e
ob
s
e
r
ve
d
m
a
iz
e
yi
e
ld
a
nd
th
e
pr
e
di
c
ti
ons
.
M
A
E
m
e
a
s
ur
e
s
th
e
pe
r
c
e
nt
a
ge
of
th
e
m
e
a
n
de
vi
a
ti
on
of
th
e
pr
e
di
c
te
d
m
a
i
z
e
yi
e
ld
f
r
om
obs
e
r
va
ti
ons
,
w
hi
le
R
M
S
E
m
e
a
s
ur
e
s
th
e
de
vi
a
ti
on
of
th
e
pr
e
di
c
te
d
m
a
iz
e
yi
e
ld
f
r
om
obs
e
r
va
ti
ons
.
M
S
E
m
e
a
s
ur
e
s
th
e
s
qua
r
e
of
th
e
d
e
vi
a
ti
on
f
r
om
obs
e
r
va
ti
ons
. T
h
e
de
ta
il
e
d f
or
m
ul
a
s
a
r
e
de
f
in
e
d a
s
f
ol
lo
w
s
:
2
=
(
∑
(
−
̅
)
(
−
̅
=
1
)
)
2
∑
(
−
̅
)
2
(
−
̅
)
2
=
1
(
1)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
. 13, No. 4, D
e
c
e
m
be
r
2024:
3930
-
3941
3936
=
√
1
∑
(
−
)
2
=
1
(
2)
=
1
∑
|
−
|
=
1
(
3)
=
1
∑
(
−
)
2
=
1
(
4)
is
t
he
obs
e
r
ve
d m
a
iz
e
yi
e
ld
,
̅
is
t
he
c
or
r
e
s
ponding m
e
a
n va
lu
e
,
n i
s
th
e
numbe
r
of
da
ta
poi
nt
s
f
or
th
e
ML
m
ode
l.
f
i
is
t
he
pr
e
di
c
te
d m
a
iz
e
yi
e
ld
,
̅
is
t
he
c
or
r
e
s
ponding m
e
a
n va
lu
e
, a
nd
p
is
th
e
t
ot
a
l
num
be
r
o
f
e
xpl
a
na
to
r
y
va
r
ia
bl
e
s
in
th
e
m
ode
l.
T
he
c
lo
s
e
r
R
2
is
to
1
or
th
e
c
lo
s
e
r
M
S
E
,
M
A
E
,
a
nd
R
M
S
E
a
r
e
to
0,
th
e
be
tt
e
r
t
he
m
ode
l’
s
pr
e
di
c
ti
on pe
r
f
or
m
a
nc
e
. A
ll
f
our
m
e
tr
ic
s
w
e
r
e
us
e
d t
o c
hoos
e
t
he
be
s
t
m
ode
l.
2.3. P
r
e
d
ic
t
io
n
of
m
ai
z
e
yi
e
ld
s
w
it
h
u
n
s
u
p
e
r
vi
s
e
d
m
ac
h
in
e
l
e
ar
n
in
g m
e
t
h
od
s
A
s
s
oc
ia
ti
on
r
ul
e
s
a
r
e
de
f
in
e
d
by
ML
to
e
s
ta
bl
is
h
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
va
r
ia
bl
e
s
.
T
he
F
P
G
a
lg
or
it
hm
is
us
e
d
to
e
s
ta
bl
is
h
th
e
r
ul
e
s
f
r
om
th
e
“
m
lx
te
nd
”
pa
c
ka
ge
.
F
P
G
is
a
n
im
pr
ove
d
ve
r
s
io
n
o
f
th
e
w
e
ll
-
known
A
pr
io
r
i
a
lg
or
it
hm
w
it
h
a
c
c
ur
a
te
e
xt
r
a
c
ti
on
e
f
f
ic
ie
nc
y
[
26]
.
T
o
do
th
is
,
it
s
e
ts
a
m
in
im
um
th
r
e
s
hol
d,
w
hi
c
h
is
u
s
e
d
to
r
e
m
ove
ir
r
e
le
va
nt
in
f
or
m
a
ti
on
in
th
e
d
a
ta
a
nd
f
or
m
f
r
e
que
nt
it
e
m
s
e
ts
.
T
he
s
e
l
a
te
r
a
r
e
us
e
d
to
a
s
s
e
s
s
t
he
s
im
il
a
r
it
y de
gr
e
e
of
di
f
f
e
r
e
nt
va
r
ia
bl
e
s
a
nd e
s
ta
bl
is
h t
he
a
s
s
o
c
ia
ti
on r
ul
e
s
.
I
nput
a
nd
out
put
va
r
ia
bl
e
s
w
e
r
e
e
n
c
ode
d
in
lo
w
,
m
e
di
um
,
a
n
d
hi
gh
ba
s
e
d
on
th
e
f
ix
e
d
th
r
e
s
hol
d
va
lu
e
s
.
M
in
im
um
s
uppor
t
w
a
s
f
ix
e
d a
s
0.25.
A
s
f
or
th
e
th
r
e
s
ho
ld
c
onf
id
e
nc
e
,
it
w
a
s
s
e
t
to
0.75.
T
h
e
r
ul
e
s
a
r
e
th
e
n
or
de
r
e
d
a
nd
s
c
r
e
e
ne
d
ba
s
e
d
on
c
onf
id
e
nc
e
a
nd
li
f
t.
A
f
te
r
f
il
te
r
in
g,
onl
y
th
e
be
s
t
r
ul
e
s
f
r
om
e
a
c
h
a
gr
o
-
c
li
m
a
ti
c
z
one
w
e
r
e
r
e
ta
in
e
d.
2.3.1. Cr
it
e
r
ia
f
or
c
h
oo
s
in
g t
h
e
b
e
s
t
as
s
oc
ia
t
io
n
r
u
le
s
D
if
f
e
r
e
nt
m
e
tr
ic
s
w
e
r
e
us
e
d
to
obt
a
in
th
e
be
s
t
r
ul
e
s
,
s
uc
h
a
s
s
uppor
t,
c
onf
id
e
nc
e
,
a
nd
li
f
t.
T
he
s
uppor
t
di
s
pl
a
ys
th
e
pr
opor
ti
on
of
tr
a
ns
a
c
ti
ons
,
in
c
lu
di
ng
it
e
m
s
X
a
nd
Y
.
T
he
c
onf
id
e
nc
e
e
xpr
e
s
s
e
s
how
m
a
ny
tr
a
ns
a
c
ti
ons
ha
ve
X
a
nd
Y
.
A
t
th
e
s
a
m
e
ti
m
e
,
th
e
li
f
t
a
s
s
e
s
s
e
s
th
e
ir
li
nk.
n
r
e
pr
e
s
e
nt
s
th
e
to
ta
l
num
be
r
of
tr
a
ns
a
c
ti
ons
,
n(
X
∪
Y
)
is
th
e
num
be
r
of
tr
a
n
s
a
c
ti
ons
w
it
h
X
or
Y
,
a
nd
n(
X
)
is
th
e
num
be
r
of
tr
a
n
s
a
c
ti
ons
w
it
h X
. A
r
ul
e
i
s
us
e
f
ul
w
he
n t
he
s
uppor
t
a
nd
c
onf
id
e
nc
e
a
r
e
hi
ghe
r
t
ha
n t
he
m
in
im
um
s
uppor
t
a
nd m
in
im
um
c
onf
id
e
nc
e
t
hr
e
s
hol
d, a
nd l
if
t
is
gr
e
a
te
r
or
e
qua
l
to
1.
(
→
)
=
(
,
)
=
(
∪
)
(
5)
(
→
)
=
(
∪
)
(
)
(
6)
(
→
)
=
(
,
)
(
)
(
)
(
7)
3.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
3.1. M
ai
z
e
yi
e
ld
m
od
e
li
n
g
T
a
bl
e
2
pr
e
s
e
nt
s
th
e
pe
r
f
or
m
a
nc
e
of
th
e
m
ode
ls
us
in
g
c
om
pa
r
i
s
on
m
e
tr
ic
s
.
B
a
s
e
d
on
th
e
r
e
s
ul
ts
,
a
ll
m
ode
ls
ha
d
a
n
M
S
E
va
lu
e
le
s
s
th
a
n
0
.
03
kg/
ha
.
T
he
D
T
te
c
hni
que
r
e
c
or
de
d
th
e
lo
w
e
s
t
M
A
E
a
nd
R
M
S
E
va
lu
e
s
(
0.0008
kg/
ha
a
nd
0.012
kg/
ha
,
r
e
s
pe
c
ti
ve
ly
)
.
T
hi
s
c
onc
lu
s
io
n
w
a
s
c
onf
ir
m
e
d
by
V
e
e
na
dha
r
i
e
t
al
.
[
27]
in
e
s
ta
bl
is
hi
ng
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
c
li
m
a
te
pa
r
a
m
e
te
r
s
a
nd
s
oybe
a
n
yi
e
ld
.
A
s
hw
in
ir
a
ni
a
nd
V
id
ya
va
th
i
[
28]
a
ppl
ie
d t
h
is
m
e
th
od t
o
im
pr
ove
t
he
m
e
th
odol
o
gy f
or
m
ode
li
ng s
uga
r
c
a
ne
yi
e
ld
. A
c
c
or
di
ng t
o
F
a
n
e
t
al
.
[
29]
,
D
T
ha
s
th
e
a
bi
li
ty
not
onl
y
to
c
a
pt
ur
e
non
-
li
ne
a
r
r
e
la
ti
ons
hi
ps
but
a
ls
o
to
c
a
pt
ur
e
e
xt
r
e
m
e
c
a
s
e
s
. A
m
ong the
di
f
f
e
r
e
nt
s
upe
r
vi
s
e
d m
e
th
ods
a
ppl
ie
d, A
N
N
a
nd R
N
N
w
e
r
e
f
ound to be the
l
e
a
s
t
pr
e
di
c
ti
ve
.
A
N
N
a
nd
R
N
N
a
r
e
m
or
e
s
e
ns
it
iv
e
to
s
a
m
pl
e
s
iz
e
,
w
hi
c
h
i
s
r
e
la
ti
ve
ly
lo
w
in
our
s
tu
dy.
K
N
N
a
ls
o
ha
s
be
tt
e
r
a
c
c
ur
a
c
y
th
a
n
th
e
ot
he
r
m
ode
ls
,
e
xc
e
pt
th
e
D
T
.
T
hi
s
r
e
s
ul
t
is
c
or
r
obor
a
te
d
by
S
ha
koor
e
t
al
.
[
30
]
,
w
ho
r
e
ve
a
le
d
th
e
s
upe
r
io
r
it
y
of
D
T
c
om
pa
r
e
d
to
K
N
N
r
e
ga
r
di
ng
pr
e
di
c
ti
on
e
r
r
or
pe
r
c
e
nt
a
ge
.
F
ur
th
e
r
m
or
e
,
th
e
r
e
s
e
a
r
c
h
by
J
a
ha
n
a
nd
S
ha
ha
r
ia
r
[
31]
us
e
d
a
DT
a
lg
or
it
hm
to
pr
e
di
c
t
f
e
r
ti
li
z
e
r
tr
e
a
tm
e
nt
in
m
a
iz
e
c
ul
ti
va
ti
on
a
nd
obs
e
r
ve
d
good
a
c
c
ur
a
c
y.
R
e
ye
s
e
t
al
.
[
32]
s
how
e
d
th
a
t
K
N
N
w
a
s
a
pr
om
is
in
g
m
e
th
od
f
or
pr
e
di
c
ti
ng
w
in
te
r
gr
a
in
yi
e
ld
s
.
F
ig
ur
e
4
il
lu
s
tr
a
te
s
a
li
ne
a
r
r
e
la
ti
ons
hi
p
in
a
ll
gr
a
phs
,
s
ugge
s
ti
ng
li
tt
le
di
f
f
e
r
e
nc
e
be
twe
e
n
th
e
a
c
tu
a
l
a
nd
pr
e
di
c
te
d
m
a
iz
e
yi
e
ld
.
T
he
pr
e
di
c
t
e
d
va
r
ia
bl
e
of
K
N
N
a
n
d
S
V
M
m
ode
ls
is
not
f
ul
ly
c
or
r
e
la
te
d
w
it
h
th
e
a
c
tu
a
l
va
r
ia
bl
e
s
how
n
in
F
ig
ur
e
s
4(
a
)
a
nd
4(
b)
.
T
he
D
T
s
ho
w
s
a
pe
r
f
e
c
t
c
or
r
e
la
ti
on
be
twe
e
n
a
c
tu
a
l
a
nd
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
U
s
in
g patt
e
r
n m
in
in
g t
o de
te
r
m
in
e
f
in
e
c
li
m
at
ic
par
am
e
te
r
s
f
or
m
ai
z
e
y
ie
ld
…
(
Souand P
e
ac
e
G
lo
r
ia
T
ahi
)
3937
pr
e
di
c
te
d
va
r
ia
bl
e
s
s
how
n
in
F
ig
ur
e
4
(
c
)
.
M
or
e
ove
r
,
th
e
pr
e
di
c
te
d
va
r
ia
bl
e
of
bot
h
A
N
N
a
nd
R
N
N
m
ode
ls
is
not
e
nt
ir
e
ly
c
or
r
e
la
te
d
w
it
h
th
e
a
c
tu
a
l
va
r
ia
bl
e
,
a
s
s
how
n
in
F
ig
ur
e
s
4(
d)
a
nd
4(
e
)
.
T
he
c
or
r
e
la
ti
on
c
oe
f
f
ic
ie
nt
(
R
2
)
f
or
e
a
c
h
m
ode
l
in
di
c
a
te
s
th
e
m
ode
l’
s
th
r
e
s
hol
d
of
pr
e
d
ic
ti
ng
m
a
iz
e
yi
e
ld
f
r
om
w
e
a
th
e
r
pa
r
a
m
e
te
r
s
.
H
ow
e
ve
r
,
a
hi
ghe
r
R
2
m
e
a
ns
th
a
t
th
e
s
tr
e
ngt
h
of
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
th
e
yi
e
ld
a
nd
c
li
m
a
t
e
va
r
ia
bl
e
s
is
c
onvi
nc
in
g.
A
ll
f
iv
e
m
ode
l
s
us
e
d
to
pr
e
di
c
t
m
a
iz
e
yi
e
ld
ha
v
e
R
2
va
lu
e
s
hi
ghe
r
th
a
n
0.60.
D
T
r
e
c
or
de
d
th
e
hi
ghe
s
t
R
2
w
hi
c
h i
s
gr
e
a
te
r
t
ha
n 0.95. C
on
s
e
que
nt
ly
, a
ll
of
t
he
d
is
c
us
s
io
n
s
a
bove
c
onf
ir
m
e
d t
ha
t
th
e
D
T
m
ode
l
pe
r
f
or
m
s
w
e
ll
c
om
pa
r
e
d t
o ot
he
r
m
ode
ls
.
T
a
bl
e
2.
R
M
S
E
,
M
A
E
,
R
2
, a
nd
M
S
E
of
K
N
N
, S
V
M
, A
N
N
, R
N
N
, a
nd D
T
M
ode
l
s
R
M
S
E
M
A
E
R
2
M
S
E
KNN
0.132
0.098
0.755
0.017
S
V
M
0.145
0.110
0.704
0.021
ANN
0.16
0.125
0.629
0.028
R
N
N
0.156
0.116
0.664
0.023
DT
0.012
0.0008
0.998
0.021
(
a
)
(
b)
(
c
)
(
d)
(
e
)
F
ig
ur
e
4. S
c
a
tt
e
r
pl
ot
s
of
a
c
tu
a
l
out
put
ve
r
s
us
pr
e
di
c
te
d output
us
in
g di
f
f
e
r
e
nt
M
L
te
c
hni
que
s
:
(
a
)
KNN
a
lg
or
it
hm
, (
b)
S
V
M
, (
c
)
DT
, (
d)
ANN
,
a
nd
(
e
)
R
N
N
. D
a
s
he
d l
in
e
:
r
e
gr
e
s
s
io
n l
in
e
,
bl
ue
l
in
e
s
:
e
r
r
or
ba
r
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
. 13, No. 4, D
e
c
e
m
be
r
2024:
3930
-
3941
3938
3.2. As
s
oc
ia
t
io
n
r
u
le
s
N
e
w
da
ta
w
e
r
e
ge
ne
r
a
te
d
us
in
g
th
e
D
T
m
ode
l
to
ha
ve
c
r
e
di
b
le
r
ul
e
s
to
id
e
nt
if
y
e
a
r
ly
m
a
iz
e
yi
e
ld
f
r
om
w
e
a
th
e
r
c
ha
r
a
c
te
r
is
ti
c
s
,
10,000
f
or
e
a
c
h
di
s
tr
ic
t.
T
he
F
P
G
a
lg
or
i
th
m
is
a
ppl
ie
d
to
th
e
da
ta
.
S
in
c
e
F
P
G
us
e
s
i
te
m
s
, t
he
nume
r
ic
a
l
a
tt
r
ib
ut
e
va
lu
e
s
a
r
e
c
onve
r
te
d i
nt
o or
di
na
l
a
tt
r
ib
ut
e
s
(
lo
w
, m
e
di
um
, a
nd high s
c
or
e
s
)
ba
s
e
d
on
f
ix
e
d
th
r
e
s
hol
d
va
lu
e
s
.
M
in
im
um
s
uppor
t
a
nd
m
in
i
m
um
c
onf
id
e
nc
e
w
e
r
e
,
r
e
s
pe
c
ti
ve
ly
,
0.25
a
nd
0.75.
T
he
s
e
va
lu
e
s
ha
ve
be
e
n
c
ho
s
e
n
not
to
o
s
m
a
ll
or
to
o
la
r
g
e
to
a
ll
ow
th
e
F
P
G
a
lg
o
r
it
hm
to
obt
a
in
a
ll
th
e
a
va
il
a
bl
e
e
le
m
e
nt
s
f
r
om
th
e
da
ta
ba
s
e
w
it
hout
unde
r
or
ove
r
e
s
ti
m
a
ti
on.
T
he
unf
il
te
r
e
d
li
s
t
of
th
e
r
ul
e
s
w
a
s
obt
a
in
e
d
a
f
te
r
m
ode
li
ng.
I
t
w
a
s
th
e
n
s
or
te
d
by
li
f
t
a
nd
c
onf
id
e
nc
e
.
O
nl
y
th
e
be
s
t
r
ul
e
s
w
e
r
e
c
hos
e
n
f
r
om
th
is
ne
w
r
ul
e
li
s
t
in
e
a
c
h
a
gr
o
-
c
li
m
a
ti
c
z
one
.
A
c
c
or
di
ng
to
th
e
r
ul
e
s
e
le
c
ti
on
c
r
it
e
r
ia
de
s
c
r
ib
e
d
a
bove
,
th
e
s
tr
ong
a
s
s
oc
ia
ti
on
r
ul
e
s
w
e
r
e
s
e
le
c
te
d
in
e
a
c
h
a
gr
o
-
c
li
m
a
ti
c
z
one
.
T
he
li
s
t
of
th
e
tr
a
in
e
d
r
ul
e
s
ha
d
a
t
le
a
s
t
th
r
e
e
a
tt
r
ib
ut
e
s
f
or
a
nt
e
c
e
de
nt
s
of
e
a
c
h
of
th
e
r
ul
e
s
f
or
m
e
d.
T
he
b
e
s
t
r
ul
e
s
w
e
r
e
c
ho
s
e
n
f
r
om
e
a
c
h
di
s
tr
ic
t
in
bot
h
a
gr
o
-
c
li
m
a
ti
c
. T
he
a
s
s
o
c
ia
ti
on r
ul
e
s
l
is
te
d i
n T
a
bl
e
s
3 a
nd 4 s
ug
ge
s
t
th
e
m
os
t
s
ui
ta
bl
e
c
ondi
ti
ons
f
or
obt
a
in
in
g
hi
gh
m
a
iz
e
yi
e
ld
.
I
n T
a
bl
e
s
3
a
nd
4,
th
e
s
uppor
t
va
lu
e
s
in
di
c
a
te
th
e
in
vol
ve
m
e
nt
of
a
gi
v
e
n
s
e
t
of
it
e
m
s
in
th
e
a
nt
e
c
e
de
nt
of
a
s
s
o
c
ia
ti
on r
ul
e
s
. T
he
c
onf
id
e
nc
e
va
lu
e
e
xpr
e
s
s
e
s
s
uppor
t
f
or
oc
c
ur
r
e
nc
e
s
of
t
r
a
ns
a
c
ti
ons
, w
hi
le
li
f
t
s
how
s
a
n i
nde
pe
nde
nt
r
e
la
ti
ons
hi
p be
tw
e
e
n a
nt
e
c
e
de
nt
a
nd
c
ons
e
que
nt
.
T
a
bl
e
3. A
s
s
oc
ia
ti
on r
ul
e
s
f
or
w
e
a
th
e
r
a
nd ma
i
z
e
yi
e
ld
i
n S
uda
no
-
G
ui
ne
a
n z
one
obt
a
in
e
d us
in
g s
im
ul
a
te
d
da
ta
f
or
B
ohi
c
on, Ka
ndi
, S
a
ve
, P
a
r
a
kou, a
nd N
a
ti
ti
ngou dis
tr
ic
t
s
A
gr
o c
l
i
m
a
t
i
c
z
one
s
A
nt
e
c
e
de
nt
s
C
ons
e
que
nt
s
S
uppor
t
C
onf
i
de
nc
e
L
i
f
t
S
uda
no
-
gui
ne
a
n
B
ohi
c
on
B
ohi
c
on
T
m
i
n m
i
ddl
e
, R
R
m
i
ddl
e
, E
T
P
m
i
ddl
e
Y
i
e
l
d hi
gh
0.444
1
1
T
m
i
n m
i
ddl
e
, R
R
m
i
ddl
e
, S
un m
i
ddl
e
Y
i
e
l
d hi
gh
0.349
1
1
E
T
P
m
i
ddl
e
, R
R
m
i
ddl
e
, U
m
a
x m
i
ddl
e
, S
un m
i
ddl
e
Y
i
e
l
d hi
gh
0.259
1
1
R
R
m
i
ddl
e
, U
m
a
x m
i
ddl
e
, S
un m
i
ddl
e
Y
i
e
l
d hi
gh
0.32
1
1
E
T
P
m
i
ddl
e
, U
m
a
x m
i
ddl
e
, S
un m
i
ddl
e
, T
m
a
x m
i
ddl
e
Y
i
e
l
d hi
gh
0.258
1
1
S
a
vè
T
m
i
n m
i
ddl
e
, U
m
a
x m
i
ddl
e
, T
m
a
x m
i
ddl
e
Y
i
e
l
d hi
gh
0.411
1
1
T
m
i
n m
i
ddl
e
, E
T
P
hi
gh, T
m
a
x m
i
ddl
e
Y
i
e
l
d hi
gh
0.357
1
1
T
m
i
n m
i
ddl
e
, R
R
m
i
ddl
e
, U
m
a
x m
i
ddl
e
Y
i
e
l
d hi
gh
0.392
1
1
T
m
i
n m
i
ddl
e
, R
R
m
i
ddl
e
, U
m
a
x m
i
ddl
e
, T
m
a
x m
i
ddl
e
Y
i
e
l
d hi
gh
0.269
1
1
R
R
m
i
ddl
e
, U
m
a
x m
i
ddl
e
, T
m
a
x m
i
ddl
e
Y
i
e
l
d hi
gh
0.339
1
1
T
a
bl
e
4. A
s
s
oc
ia
ti
on r
ul
e
s
f
or
w
e
a
th
e
r
a
nd ma
i
z
e
yi
e
ld
i
n S
uda
ni
a
n z
one
obt
a
in
e
d us
in
g
s
im
ul
a
te
d da
ta
f
or
B
ohi
c
on, Ka
ndi
, S
a
ve
, P
a
r
a
kou, a
nd N
a
ti
ti
ngou dis
tr
ic
ts
A
gro
-
c
l
i
m
a
t
i
c
z
one
s
A
nt
e
c
e
de
n
t
s
Cons
e
que
nt
s
S
upport
Confi
de
nc
e
L
i
ft
S
uda
ni
a
n
z
one
P
a
ra
kou
T
m
i
n
m
i
dd
l
e
,
U
m
a
x
m
i
ddl
e
,
T
m
a
x
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
513
1
1
E
T
P
m
i
ddl
e
,
S
un
m
i
ddl
e
,
T
m
a
x
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
391
1
1
T
m
i
n
m
i
dd
l
e
,
E
T
P
m
i
ddl
e
,
S
un
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
385
1
1
Tmin
m
i
dd
l
e
,
E
T
P
m
i
ddl
e
,
S
un
m
i
ddl
e
,
T
m
a
x
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
319
1
1
T
m
i
n
m
i
dd
l
e
,
E
T
P
m
i
ddl
e
,
U
m
a
x
m
i
ddl
e
,
S
un
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
311
1
1
K
a
ndi
E
T
P
m
i
ddl
e
,
U
m
a
x
m
i
ddl
e
,
T
m
a
x
m
i
dd
l
e
Y
i
e
l
d
hi
gh
0.
616
1
1
RR
m
i
ddl
e
,
U
m
a
x
m
i
ddl
e
,
S
un
m
i
ddl
e
,
T
m
a
x
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
324
1
1
T
m
i
n
m
i
dd
l
e
,
RR
m
i
ddl
e
,
T
m
a
x
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
417
1
1
T
m
i
n
m
i
dd
l
e
,
RR
m
i
ddl
e
,
U
m
a
x
m
i
ddl
e
,
T
m
a
x
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
374
1
1
T
m
i
n
m
i
dd
l
e
,
RR
m
i
ddl
e
,
E
T
P
m
i
ddl
e
,
T
m
a
x
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
361
1
1
N
a
t
i
t
i
ngou
U
m
a
x
l
ow
,
E
T
P
m
i
ddl
e
,
T
m
i
n
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
605
1
1
U
m
a
x
l
ow
,
E
T
P
m
i
ddl
e
,
T
m
i
n
m
i
ddl
e
,
S
un
m
i
ddl
e
,
RR
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
294
1
1
RR
m
i
ddl
e
,
E
T
P
m
i
ddl
e
,
S
un
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
475
1
1
Tmin
m
i
dd
l
e
,
RR
m
i
ddl
e
,
S
un
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
376
1
1
T
m
i
n
m
i
dd
l
e
,
RR
m
i
ddl
e
,
E
T
P
m
i
ddl
e
,
S
un
m
i
ddl
e
Y
i
e
l
d
hi
gh
0.
375
1
1
T
he
li
s
t
of
r
ul
e
s
d
e
f
in
e
s
a
s
s
oc
ia
ti
on
s
be
twe
e
n
th
e
m
e
di
um
va
lu
e
of
th
e
m
in
im
um
a
nd
th
e
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
,
hum
id
it
y,
s
uns
tr
oke
,
r
a
in
f
a
ll
,
a
nd
e
v
a
pot
r
a
ns
pi
r
a
t
io
n.
F
or
e
xa
m
pl
e
,
th
e
be
s
t
m
a
iz
e
tr
e
nds
w
e
r
e
obs
e
r
ve
d
in
th
e
S
uda
no
-
G
ui
ne
a
n
z
one
f
or
m
e
di
um
va
lu
e
s
of
m
in
im
um
te
m
pe
r
a
tu
r
e
,
r
a
in
f
a
ll
,
a
nd
e
va
pot
r
a
ns
pi
r
a
ti
on
(
E
T
P
=
[
2
.
513−
3
.
628
mm
[
;
RR
=
[
2
.
314−
6
.
767
mm
[
;
T
m
in
=
[
22
.
088−
23
.
88
°
C
[)
.
M
a
iz
e
yi
e
ld
w
a
s
a
ls
o
hi
gh
w
he
n
th
e
m
in
im
um
te
m
pe
r
a
tu
r
e
,
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
,
a
nd
m
a
xi
m
um
hum
id
it
y
w
e
r
e
m
e
di
um
(
T
m
in
=
[
22
.
088−
23
.
88
°
C
[
;
T
m
ax
=
[
29
.
836−
33
.
599
◦
C
[
;
U
m
ax
=
[
93
.
497−
96
.
302%
[
)
.
T
hi
s
m
e
a
n
s
th
a
t
m
a
iz
e
yi
e
ld
i
s
hi
gh
in
th
is
a
gr
o
-
c
li
m
a
ti
c
z
one
w
he
n
th
e
s
e
c
li
m
a
te
a
tt
r
ib
ut
e
s
a
r
e
m
e
di
um
ove
r
th
e
pr
oduc
ti
on
pe
r
io
d.
I
n
th
e
S
uda
ni
a
n
z
one
,
th
e
yi
e
ld
w
a
s
hi
gh
w
he
n
e
va
pot
r
a
ns
pi
r
a
ti
on,
m
a
xi
m
um
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
U
s
in
g patt
e
r
n m
in
in
g t
o de
te
r
m
in
e
f
in
e
c
li
m
at
ic
par
am
e
te
r
s
f
or
m
ai
z
e
y
ie
ld
…
(
Souand P
e
ac
e
G
lo
r
ia
T
ahi
)
3939
hum
id
it
y,
a
nd
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
w
e
r
e
m
e
di
um
(
E
T
P
=
[
1
.
618−
4
.
801
mm
[
;
U
m
ax
=
[
87
.
324−
102
.
28
%
[
;
T
m
ax
=
[
29
.
371−
35
.
971
°
C
[)
o
r
w
he
n
e
va
pot
r
a
ns
pi
r
a
ti
on
a
nd
m
in
im
um
te
m
pe
r
a
tu
r
e
w
e
r
e
m
e
di
um
a
nd
m
a
xi
m
um
hum
id
it
y
w
a
s
lo
w
(
E
T
P
=
[
1
.
618−
4
.
801
mm
[
;
U
m
ax
=
<
87
.
324%
;
T
m
in
=
[
21
.
337−
24
.
015
◦
C
[)
.
H
ow
e
ve
r
,
w
he
n
th
e
va
lu
e
s
of
m
in
im
um
te
m
pe
r
a
tu
r
e
,
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
,
a
nd
m
a
xi
m
um
hum
id
it
y
w
e
r
e
m
e
di
um
,
th
e
m
a
iz
e
yi
e
ld
w
a
s
a
l
s
o
hi
gh
in
th
e
S
uda
ni
a
n
z
one
(
T
m
ax
=
[
29
.
371−
35
.
971
°
C
[
;
U
m
ax
=
[
87
.
324−
102
.
28%
[
;
T
m
in
=
[
21
.
337−
24
.
015
°
C
[)
.
A
pa
r
t
f
r
om
a
lo
w
hum
id
it
y
obt
a
in
e
d
in
th
e
S
ouda
ni
a
n
z
one
,
non
e
of
th
e
obt
a
in
e
d
r
ul
e
s
ha
d
a
lo
w
a
tt
r
ib
ut
e
.
A
s
im
il
a
r
s
tu
dy
K
a
ur
a
nd
A
tt
w
a
l
[
13]
on
th
e
e
f
f
e
c
t
of
te
m
pe
r
a
tu
r
e
a
nd
r
a
in
f
a
ll
on
r
ic
e
yi
e
ld
in
P
unj
a
b
c
it
y
c
or
r
obor
a
te
s
th
e
r
e
s
ul
t
s
.
T
he
y
c
onc
lu
de
d
th
a
t
hi
gh
r
ic
e
yi
e
ld
is
obs
e
r
ve
d
w
he
n
th
e
te
m
pe
r
a
tu
r
e
is
m
e
di
um
d
ur
in
g
th
e
ve
ge
ta
ti
ve
pha
s
e
or
w
he
n
r
a
in
f
a
ll
is
lo
w
dur
in
g
th
e
gr
a
in
f
il
li
ng
a
nd
m
a
tu
r
it
y
pha
s
e
.
S
upr
o'
s
[
15]
w
or
k
on
r
ic
e
in
I
ndi
a
s
uppor
ts
th
e
r
e
s
ul
ts
.
H
e
s
how
e
d
th
a
t
r
ic
e
yi
e
ld
i
s
hi
gh
und
e
r
hi
gh
te
m
pe
r
a
tu
r
e
s
a
nd
m
e
d
iu
m
hum
id
it
y.
R
a
o
e
t
al
.
[
1]
e
s
ti
m
a
ti
ng
p
a
ddy
yi
e
ld
pr
e
di
c
to
r
s
us
in
g
te
m
pe
r
a
tu
r
e
,
r
a
in
f
a
ll
,
s
oi
l
pH
,
a
nd
ni
tr
og
e
n,
obs
e
r
ve
d
th
a
t
hi
gh
yi
e
ld
s
w
e
r
e
a
s
s
oc
ia
te
d
w
it
h
hi
gh
te
m
pe
r
a
tu
r
e
,
r
a
in
f
a
ll
,
a
nd
m
e
di
um
pH
a
nd
ni
tr
oge
n.
N
o
r
ul
e
s
f
ound
by
th
e
s
e
a
ut
hor
s
a
r
e
de
ve
lo
pe
d
w
it
h
lo
w
a
tt
r
ib
ut
e
s
.
C
ir
a
d'
s
[
33]
s
tu
dy
hi
ghl
ig
ht
s
c
li
m
a
te
c
ha
n
ge
'
s
s
ig
ni
f
ic
a
nt
im
pa
c
t
on
m
a
iz
e
yi
e
ld
s
in
s
ub
-
S
a
ha
r
a
n A
f
r
ic
a
. I
t
e
m
pha
s
iz
e
s
t
ha
t
a
4
°C
t
e
m
pe
r
a
tu
r
e
i
nc
r
e
a
s
e
c
oul
d de
c
r
e
a
s
e
m
a
iz
e
yi
e
ld
s
by a
ppr
oxi
m
a
te
ly
14%
.
F
ur
th
e
r
m
or
e
,
th
e
s
tu
dy
s
ugg
e
s
ts
th
a
t
a
d
a
pt
a
ti
on
of
f
e
r
ti
li
z
a
ti
on
is
ne
c
e
s
s
a
r
y
f
or
m
a
iz
e
c
ul
ti
va
ti
on,
e
s
pe
c
ia
ll
y
in
th
e
c
a
s
e
of
s
ubs
ta
nt
ia
l
te
m
pe
r
a
tu
r
e
in
c
r
e
a
s
e
s
.
I
nde
e
d,
r
e
s
e
a
r
c
h
by
G
una
th
il
a
ke
e
t
al
.
[
34]
r
e
ve
a
le
d
th
a
t
in
c
r
e
a
s
e
s
in
m
in
im
um
te
m
pe
r
a
tu
r
e
a
r
e
g
e
ne
r
a
ll
y
m
or
e
s
ig
ni
f
ic
a
nt
th
a
n
in
c
r
e
a
s
e
s
in
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
.
T
he
s
tu
dy
by
K
ouda
he
e
t
al
.
[
35]
on
th
e
im
pa
c
t
of
c
li
m
a
te
va
r
ia
bi
li
ty
on c
r
op
yi
e
ld
s
in
S
out
he
r
n
T
ogo
c
onc
lu
de
d
th
a
t
te
m
pe
r
a
tu
r
e
in
c
r
e
a
s
e
s
ig
ni
f
ic
a
nt
ly
a
f
f
e
c
ts
th
e
yi
e
ld
s
of
m
a
iz
e
a
nd
be
a
ns
.
P
r
e
c
ip
it
a
ti
on
a
ls
o
r
e
m
a
in
s
a
f
unda
m
e
nt
a
l
phe
nom
e
non
of
th
e
c
li
m
a
te
s
y
s
te
m
,
c
ons
id
e
r
a
bl
y
im
pa
c
ti
ng
a
ll
a
s
pe
c
t
s
of
li
f
e
,
in
c
lu
di
ng
a
gr
ic
ul
tu
r
e
[
36]
.
H
ow
e
ve
r
,
m
a
iz
e
w
a
s
a
ls
o s
how
n
to
be
le
s
s
r
e
a
c
ti
ve
to
r
is
in
g
a
tm
o
s
phe
r
ic
C
O
2
a
nd
le
s
s
a
f
f
e
c
te
d
by
hi
gh
e
r
te
m
pe
r
a
tu
r
e
s
or
lo
w
e
r
r
a
in
f
a
ll
w
it
hout
n
it
r
oge
n
f
e
r
ti
li
z
a
ti
on
[
37]
.
B
ut
it
be
c
om
e
s
m
or
e
vul
ne
r
a
bl
e
to
ni
tr
oge
n
de
f
ic
ie
nc
y,
w
it
h
m
o
r
e
s
ig
ni
f
ic
a
nt
n
e
ga
ti
ve
e
f
f
e
c
ts
on
yi
e
ld
.
T
he
s
e
c
onc
lu
s
io
ns
unde
r
s
c
or
e
th
e
c
r
uc
ia
l
im
por
ta
nc
e
of
a
da
pt
in
g
c
r
op
f
e
r
ti
li
z
a
ti
on,
pa
r
ti
c
ul
a
r
ly
f
or
m
a
iz
e
,
in
th
e
f
a
c
e
of
th
e
c
ha
ll
e
nge
s
of
c
li
m
a
te
c
ha
nge
in
s
ub
-
S
a
ha
r
a
n
A
f
r
ic
a
.
T
he
r
e
f
o
r
e
,
it
is
im
pe
r
a
ti
ve
to
im
pl
e
m
e
nt
s
us
ta
in
a
bl
e
w
a
te
r
m
a
na
ge
m
e
nt
a
nd
c
li
m
a
te
c
ha
nge
a
da
pt
a
ti
on
s
tr
a
te
gi
e
s
th
a
t
c
a
n
he
lp
im
pr
ove
c
r
op
pe
r
f
or
m
a
nc
e
in
s
ub
-
S
a
ha
r
a
n
A
f
r
ic
a
.
T
he
r
e
s
ul
ts
of
th
is
s
tu
dy
hi
ghl
ig
ht
th
e
e
s
s
e
nt
ia
l
c
ondi
ti
ons
f
or
a
c
hi
e
vi
ng
opt
im
a
l
m
a
iz
e
yi
e
ld
s
.
T
o
our
knowle
dge
,
no
pr
e
di
c
ti
on
of
m
a
iz
e
yi
e
ld
ha
s
be
e
n
m
a
de
in
B
e
ni
n
us
in
g
a
s
s
oc
ia
ti
on
r
ul
e
s
.
D
e
s
pi
te
t
he
c
ont
r
ib
ut
io
n of
t
hi
s
s
tu
dy, s
om
e
l
im
it
a
ti
ons
s
houl
d
be
c
ons
id
e
r
e
d i
n f
ut
ur
e
w
o
r
ks
. T
hi
s
s
tu
dy only
c
ons
id
e
r
s
c
li
m
a
ti
c
f
a
c
to
r
s
a
s
in
put
va
r
ia
bl
e
s
a
nd
th
u
s
di
s
r
e
ga
r
ds
ot
he
r
f
a
c
to
r
s
s
uc
h
a
s
f
e
r
ti
li
z
e
r
s
,
pe
s
t
m
a
na
ge
m
e
nt
,
a
gr
ic
ul
tu
r
a
l
pr
a
c
ti
c
e
s
,
a
nd
va
r
ie
ti
e
s
.
T
he
s
e
f
a
c
to
r
s
c
a
n
be
c
om
bi
ne
d
w
it
h
th
e
c
li
m
a
ti
c
va
r
ia
bl
e
s
to
pr
ovi
de
be
tt
e
r
r
ul
e
s
.
4.
C
O
N
C
L
U
S
I
O
N
T
hi
s
p
a
pe
r
ha
s
id
e
nt
if
ie
d
f
in
e
c
li
m
a
ti
c
p
a
r
a
m
e
te
r
s
f
or
hi
gh
m
a
iz
e
yi
e
ld
us
in
g
p
a
tt
e
r
n
m
in
in
g
in
B
e
ni
n
(
W
e
s
t
A
f
r
ic
a
)
.
F
ir
s
tl
y,
a
n
e
xc
e
ll
e
nt
s
upe
r
vi
s
e
d
m
ode
l
(
de
c
is
io
n
tr
e
e
)
th
a
t
pr
e
di
c
t
s
m
a
iz
e
yi
e
ld
ba
s
e
d
on
w
e
a
th
e
r
pa
r
a
m
e
te
r
s
is
pr
opos
e
d
.
S
in
c
e
th
is
m
od
e
l
is
not
e
a
s
il
y
in
te
r
pr
e
ta
bl
e
,
th
e
F
P
G
a
lg
or
it
hm
is
us
e
d
to
e
s
ta
bl
is
h
th
e
r
e
la
ti
on
be
twe
e
n
w
e
a
th
e
r
a
nd
m
a
iz
e
yi
e
ld
.
T
he
r
e
s
ul
ts
s
how
e
d
th
a
t
m
os
t
of
th
e
r
ul
e
s
f
or
hi
gh
m
a
iz
e
yi
e
ld
a
r
e
a
s
s
oc
ia
te
d
w
it
h
th
e
m
e
di
um
va
lu
e
s
of
m
in
i
m
um
a
nd
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
,
hum
id
it
y,
s
uns
tr
oke
, r
a
in
f
a
ll
, a
nd e
va
pot
r
a
ns
pi
r
a
ti
on. T
he
b
e
s
t
tr
e
nds
a
r
e
obs
e
r
ve
d i
n t
he
S
uda
no
-
G
ui
ne
a
n a
nd S
uda
ni
a
n
z
one
s
f
or
m
e
di
um
va
lu
e
s
of
m
in
im
um
te
m
pe
r
a
tu
r
e
,
m
a
xi
m
um
te
m
pe
r
a
tu
r
e
,
hum
id
it
y
or
r
a
in
f
a
ll
,
a
nd
e
va
pot
r
a
ns
pi
r
a
ti
on.
T
he
s
e
id
e
nt
if
ie
d
r
ul
e
s
a
r
e
a
r
e
li
a
bl
e
a
nd
pr
om
is
in
g
ne
w
a
ppr
oa
c
h
to
im
pr
ov
in
g
m
a
iz
e
yi
e
ld
in
B
e
ni
n.
F
r
om
th
is
s
tu
dy,
f
a
r
m
e
r
s
c
a
n
im
pl
e
m
e
nt
pr
e
ve
nt
iv
e
m
e
a
s
ur
e
s
or
a
da
pt
th
e
ir
c
ul
ti
va
ti
on
pr
a
c
ti
c
e
s
t
o m
it
ig
a
te
t
he
r
is
ks
of
c
li
m
a
te
c
h
a
nge
.
A
C
K
N
O
WL
E
D
G
E
M
E
N
T
S
T
hi
s
w
or
k
w
a
s
s
uppor
te
d
by
G
e
r
m
a
n
A
c
a
de
m
ic
E
xc
ha
nge
S
e
r
vi
c
e
(
D
A
A
D
)
w
it
h
gr
a
nt
num
be
r
91786177
a
nd
S
c
hol
a
r
s
hi
p
P
r
ogr
a
m
in
A
r
ti
f
ic
ia
l
I
nt
e
ll
ig
e
nc
e
f
o
r
D
e
ve
lo
pm
e
nt
(
A
I
4D
)
A
f
r
ic
a
.
F
unde
d
by
th
e
I
nt
e
r
na
ti
ona
l
D
e
ve
lo
pm
e
nt
R
e
s
e
a
r
c
h
C
e
nt
r
e
(
I
D
R
C
)
a
nd
th
e
S
w
e
di
s
h
I
nt
e
r
na
ti
ona
l
C
oope
r
a
ti
on
A
ge
nc
y
(
S
I
D
A
)
,
a
nd
m
a
na
ge
d
by
th
e
A
f
r
ic
a
n
C
e
nt
r
e
f
or
T
e
c
hnol
ogy
S
tu
di
e
s
a
nd
D
e
ve
lo
pm
e
nt
(
A
C
T
S
)
w
it
h
gr
a
nt
num
be
r
10016012.
R
E
F
E
R
E
N
C
E
S
[
1]
P
.
R
.
R
a
o,
S
.
P
.
G
ow
da
,
a
nd
R
.
J
.
P
r
a
t
hi
bha
,
“
P
a
ddy
yi
e
l
d
pr
e
di
c
t
or
us
i
ng
t
e
m
pe
r
a
t
ur
e
,
r
a
i
nf
a
l
l
,
s
oi
l
pH
,
a
nd
ni
t
r
oge
n,”
L
e
c
t
u
r
e
N
ot
e
s
i
n E
l
e
c
t
r
i
c
al
E
ngi
ne
e
r
i
ng
, vol
. 545, pp. 245
–
253, 2019, doi
:
10.1007/
978
-
981
-
13
-
5802
-
9_23.
[
2]
H
.
H
.
K
a
nw
a
l
,
I
.
A
hm
a
d,
A
.
A
hm
a
d,
a
nd
Y
.
L
i
,
“
Y
i
e
l
d
f
or
e
c
a
s
t
i
ng
a
nd
a
s
s
e
s
s
m
e
nt
of
i
nt
e
r
a
nnua
l
w
he
a
t
yi
e
l
d
va
r
i
a
bi
l
i
t
y
us
i
ng
Evaluation Warning : The document was created with Spire.PDF for Python.