Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
12
,
No.
3
,
Decem
ber
2
01
8
, p
p.
1
0
87
~
1
0
93
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
2
.i
3
.pp
1
0
87
-
1
0
93
1087
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
Agricult
ure Data
Analyti
cs
in
Cro
p Yield
Estimati
on:
A Criti
cal Re
view
B M
Sag
ar,
C
auvery
N K
Depa
rtment
o
f
I
nform
at
ion
Sci
e
nce
& Engg, R V
Coll
eg
e
of
En
gine
er
ing, Be
ng
al
uru,
Karna
ta
k
a
,
Indi
a
Art
ic
le
In
f
o
ABST
CT
Art
ic
le
history:
Re
cei
ved
Ma
y
2
, 2
018
Re
vised Ju
n
1
5
, 2018
Accepte
d
Se
p
25, 201
8
Agric
ult
ur
e
is
important
for
hum
an
survival
bec
ause
it
serve
s
t
he
basic
nee
d
.
A
well
-
known
f
ac
t
that
th
e
m
ajorit
y
of
populati
on
(≥55%)
in
I
ndia
is
into
agr
ic
u
lt
ure
.
Due
to
var
ia
ti
ons
in
cl
imatic
condi
t
i
ons,
the
re
exi
st
bott
le
n
ec
ks
for
inc
re
asing
th
e
cro
p
p
roduc
t
io
n
in
India
.
I
t
has
bec
om
e
cha
l
le
n
ging
ta
sk
to
ac
hi
eve
desir
ed
ta
rge
ts
in
Agri
base
d
cro
p
y
iel
d.
Vari
ous
fac
to
rs
are
to
be
conside
red
whi
c
h
have
direct
i
m
pac
t
on
the
pr
oduct
ion
,
produc
ti
vi
t
y
of
the
cro
ps.
Crop
y
i
e
l
d
pre
di
ct
ion
is
one
of
the
importan
t
factors
in
agr
i
cul
tur
e
pra
ctice
s
.
Farm
ers
nee
d
infor
m
at
ion
reg
ard
in
g
cro
p
y
i
el
d
b
efo
re
sow
in
g
see
ds
in
the
ir
fi
e
lds
to
ac
hi
eve
en
hanc
ed
cro
p
y
iel
d.
Th
e
use
of
t
echnolog
y
in
agr
ic
u
lt
ure
has
i
ncr
ea
sed
in
rece
nt
y
e
ar
and
da
ta
anal
y
tics
is
on
e
such
tr
end
tha
t
has
pen
et
ra
t
ed
int
o
th
e
agr
i
c
ult
ure
fi
el
d
.
The
m
ai
n
cha
llenge
i
n
using
big
dat
a
in
ag
ric
u
lture
is
ide
nti
f
ica
ti
on
of
eff
e
ct
iv
e
ness
of
big
dat
a
ana
l
y
t
ic
s
.
Eff
orts
are
goin
g
on
to
under
stand
how
big
dat
a
anal
y
tics
ca
n
agr
ic
ul
tur
e
produc
ti
vi
t
y
.
Th
e
pre
sen
t
stud
y
give
s
insight
s
on
var
ious
data
an
aly
t
ics
m
et
hods
appl
ie
d
to
cro
p
y
i
el
d
pre
diction
and
al
so
signifies
th
e
importan
t
la
cun
ae poi
nts’
i
n
the proposed a
rea
of
rese
arc
h
.
Ke
yw
or
ds:
Agricult
ure
dat
a
Crop yi
el
d
Data analy
ti
cs
Pr
e
dicti
on
Sm
art far
m
ing
Copyright
©
201
8
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed.
Corres
pond
in
g
Aut
h
or
:
B M
Sag
a
r,
Dep
a
rtm
ent o
f Info
rm
at
ion
Sc
ie
nce &
En
gg,
R V
C
ollege
of Enginee
rin
g,
Be
ng
al
uru,
Ka
rn
at
a
ka,
India.
Em
a
il
:
sagar
bm
@r
vce.edu.I
n
1.
INTROD
U
CTION
Agricult
ure
f
orm
s
the
basis
for
f
ood
sec
uri
ty
and
he
nce
it
is
i
m
po
rtant.
In
India,
m
ajo
rity
of
th
e
popula
ti
on
i.e
.,
ab
ov
e
55%
is
dep
e
ndent
on
agr
ic
ultur
e
as
per
t
he
rece
nt
i
nfor
m
at
ion
.
A
gr
ic
ultur
e
is
t
he
fiel
d
that
ena
bles
th
e
far
m
ers
to
gr
ow
ideal
c
rops
i
n
acc
orda
nce
with
t
he
e
nv
i
r
on
m
ental
bala
nce.
I
n
I
ndia
,
wh
eat
and
rice
are
th
e
m
ajo
r
gro
w
n
crops
al
ong
with
su
ga
rca
ne,
pota
toes, o
il
seeds
et
c.
Farm
ers
al
so
grow
n
on
-
food
it
e
m
s
l
ike
rub
ber,
cotto
n,
j
ut
e
et
c.
More
th
an
70%
of
the
hous
e
hold
in
t
he
r
ur
al
area
de
pend
on
a
gr
ic
ultur
e
.
This
dom
ai
n
pr
ovides
em
plo
ym
ent
to
m
or
e
than
60%
of
the
total
popula
ti
on
an
d
has
a
con
t
rib
ution
to
GDP
al
so
(a
bout
17
%)
[1
]
.
I
n
the
f
arm
ou
tp
ut,
India
ra
nks
sec
ond
co
ns
i
der
in
g
the
world
wi
de
scena
rio
.
T
his
is
th
e
widest eco
nom
ic
sect
or
an
d h
as an
i
m
po
rtan
t ro
le
r
egardi
ng the f
ram
ewo
r
k
of
s
ocio
-
eco
no
m
ic
f
abr
ic
of I
ndia
.
Farm
ing
de
pe
nd
s
on
var
i
ous
f
act
ors
li
ke
cl
i
m
at
e
and
eco
no
m
ic
factor
s
li
ke
te
m
per
at
ur
e,
ir
r
igati
on
,
culti
vation,
so
i
l,
rain
fall
,
pest
ic
ide
and
fe
rtil
iz
ers.
Hist
or
ic
al
i
nf
orm
at
ion
reg
ar
ding
cr
op
yi
el
d
pro
vid
es
m
ajor
input
f
or
c
ompanies
en
ga
ge
d
in
this
dom
ai
n.
T
hese
co
m
pan
ie
s
m
ake
us
e
of
a
gr
ic
ul
ture
products
as
ra
w
m
at
erial
s,
anim
al
feed
,
pa
pe
r
pr
oductio
n
and
so
on.
T
he
est
im
ation
of
pro
duct
ion
of
cr
op
help
s
these
com
pan
ie
s
in
plan
ning
sup
ply
chain
decisi
on
li
ke
pro
du
c
ti
on
sche
duli
ng.
T
he
in
du
st
ries
su
c
h
as
fer
t
il
iz
ers,
seed,
a
groch
e
m
ic
al
s
and
agri
cultural
m
achi
ner
y
pla
n
pro
duct
ion
a
nd
act
i
viti
es
li
ke
m
ar
keting
base
d
on
the
est
i
m
at
es
o
f
cr
op
yi
el
d
[2
]
. Farm
ers
exp
erie
nce w
a
s the on
ly
w
ay
f
or
p
re
dicti
on
of crop y
ie
ld
in the p
ast
d
ay
s.
Tech
no
l
og
y
pe
netrati
on int
o
a
gr
ic
ultur
e
fiel
d has le
d
to
au
t
om
at
ion
of the
a
ct
ivit
ie
s li
ke
yield esti
m
a
ti
on
, c
rop
healt
h
m
on
it
ori
ng
et
c.
C
rop
y
ie
ld
pr
e
dicti
on
has
gen
e
rated
a
lot
interest
i
n
the
resea
rch
com
m
un
it
y
and
al
so
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
12
, N
o.
3
,
Dece
m
ber
2
01
8
:
1
0
87
–
1
0
93
1088
for
a
gr
ic
ultur
e
relat
ed or
gan
iz
at
ion
s.
C
rop
yi
el
d
pre
dicti
on
helps
t
he farm
ers
in
v
a
rio
us
ways b
y
pro
vi
ding th
e
record
of
pr
e
vi
ou
s
c
r
op
yi
el
d.
This
is
help
fu
l
to
gove
rn
m
ent
in
fr
am
ing
poli
ci
es
relat
ed
to
crops
s
uch
as
cr
op
insu
ran
ce p
olici
es,
supp
ly
cha
in
operati
on p
ol
ic
ie
s.
Know
i
ng
w
hat
cr
ops
ha
s
bee
n
gro
w
n,
and
how
m
uch
ar
e
a
of
it
had
bee
n
sh
ow
n
histor
ic
al
ly
,
com
bin
ed
with
the
pri
ces
at
wh
ic
h
it
c
ou
l
d
hav
e
bee
n
s
old
at
the
ne
arest
m
ark
et
-
place
provides
the i
nc
om
e
-
gr
owth
pr
of
il
e
of
t
he far
m
er [
3].
Agricult
ure
se
ct
or
is
st
rug
gling
to
i
ncr
ease
the
pro
duct
ivit
y
of
c
rop
i
n
I
ndia
.
Mo
nso
on
rainf
al
l
is
t
he
m
ai
n
so
urce
of
water
f
or
m
or
e
t
han
60
pe
rcen
t
of
t
he
crops.
Sm
art
agr
ic
ultur
e
dr
i
ven
by
I
nfo
r
m
at
ion
Tech
no
l
og
y
is
the
em
erg
ing
tren
d
i
n
the
res
earch
in
t
his
a
r
ea
in
recent
da
ys.
O
ne
of
the
areas
bein
g
e
xplo
red
is
the
pro
blem
of
yi
el
d
pr
e
dicti
on
w
hich
is
a
m
ajo
r
co
nce
rn.
Data
m
ining
te
chn
i
qu
e
s
a
re
bein
g
widely
use
d
a
s
a
par
t
of
s
olu
ti
on
for
cr
op
yi
el
d
pr
edict
io
n.
Va
rio
us
da
ta
m
ining
te
chn
i
qu
e
s
are
unde
r
eval
uatio
n
for
est
i
m
ation
of
c
rop
pro
du
ct
i
on
of
the
f
uture
y
ears
[4
]
.
Data
m
ining
is
t
he
proces
s
in
w
hic
h
the
hidden
pa
tt
ern
s
are
disco
vered
us
ing
a
naly
sis
of
la
rg
e
data
set
s.
The
data
m
ining
an
d
dat
a
analy
ti
cs
te
c
hn
i
qu
e
s
us
e
art
ific
ia
l
intel
li
gen
ce,
st
at
ist
ic
s,
m
achi
ne
le
ar
ning
an
d
data
base
sys
tem
.
In
data
m
ining,
uns
uper
vised
a
nd
supe
rv
is
e
d
m
et
ho
ds
a
re
be
ing
us
e
d.
In
un
su
pe
r
vised
le
a
r
ning,
cl
us
te
rs
a
re
f
or
m
ed
us
i
ng
la
r
ge
data
set
s
an
d
in
super
vi
sed
le
arn
in
g
cl
assif
ic
at
ion
are
done
based
on
the
data
set
s.
In
cl
us
te
rin
g
te
ch
ni
qu
e
,
‘
data
po
i
nt
s’
are
exam
ined
to
gro
up
them
into
‘clu
ste
rs’
acc
ordin
g
to
s
peci
fic
par
am
et
er.
The
data
po
i
nts
in
sam
e
c
luster
ha
ve
le
ss
di
sta
nce
com
par
ed
to
da
ta
po
ints
of
di
ff
ere
nt
cl
us
te
r
s.
The
a
naly
sis
of
the
cl
us
te
r
div
i
des
data
into
well
orga
nized
gro
up
s
. T
he na
tural str
uctu
re
of the
data is c
aptu
red by t
hes
e w
el
l
-
f
orm
ed
gro
up
s
[3, 4].
This
survey
f
oc
us
es
on
va
rio
us
m
et
ho
ds
be
ing
use
d
f
or
crop
yi
el
d
pr
e
dicti
on
.
T
he
m
e
t
hods
be
i
ng
us
e
d
are
De
nsi
ty
based
cl
ust
ering
te
c
hn
i
ques,
M
ulti
ple
Linear
regress
ion
,
Cl
ust
erin
g
la
rg
e
a
ppli
cat
ion
s
(CLAR
A)
,
Peti
ti
on
in
g
a
rou
nd
Me
do
i
ds
(P
A
M) and
de
ns
it
y base
d
cl
us
te
ri
ng alg
or
it
hm
call
ed
DBSC
A
N.
2.
REVIEW
OF
LIT
ERATUR
E MET
HO
DS OF
C
ROP YI
EL
D
PRE
DI
C
TION
At
prese
nt
we
are
at
the
im
m
ense
need
of
a
no
t
her
G
reen
r
evo
l
ution
to
s
upply
the
f
ood
dem
and
of
grow
i
ng
popul
at
ion
.
W
it
h
th
e
decr
ease
of
avail
able
culti
vab
le
la
nd
gl
obal
ly
and
the
decr
ease
d
c
ulti
vab
le
water
res
ource
s,
it
is
alm
os
t
i
m
po
ssible
to
r
eport
highe
r
cr
op
yi
el
d.
Agri
cultural
ba
sed
big
data
analy
ti
cs
is
on
e
a
ppr
oach,
belie
ved
t
o
ha
ve
a
sig
nific
ant
ro
le
a
nd
posit
ive
i
m
pact
on
the
i
ncr
ea
se
of
c
rop
yi
el
d
by
pro
vid
in
g
t
he
optim
u
m
conditi
on
f
or
t
he
pla
nt
gr
ow
t
h
a
nd
decr
easi
ng
the
yi
el
d
gaps
a
nd
the
cr
op
dam
a
ge
a
nd
wastage
.
W
it
h
this
ai
m
t
he
present
pap
e
r
re
views
ab
out
th
e
va
rio
us
a
dva
nces,
desi
gn
m
od
el
s
,
s
of
t
war
e
too
ls
and
al
gorithm
s
ap
plied
i
n
t
he
pr
e
dicti
on
assessm
ent
and
est
im
ation
of
the
cr
op
yi
el
d.
India
is
ba
sic
al
l
y
agr
ic
ultur
e
bas
ed
c
ountry
a
nd
ap
pro
xim
a
te
ly
70%
our
c
ount
ry
eco
no
m
ic
s
is
directl
y
or
i
nd
i
rectl
y
rela
te
d
to
the
ag
ricult
ural
crops
.
T
he
pr
i
nciple
c
rop
whic
h
occ
upie
s
th
e
highest
(60
-
70%)
pe
rce
ntag
e
of
culti
va
ble
la
nd
in
the
Indian
s
oil
is
the
pad
dy
culture
an
d
it
is
the
m
ajo
r
crop
especial
l
y
in
central
and
south
par
ts
of
th
e
India.
Ri
ce
crop
culti
vation
play
s
an
i
m
per
at
i
ve
par
t
in
su
ste
nan
ce
sec
ur
it
y
of
I
ndia
,
con
t
ribu
ti
ng
over 4
0%
to
gen
e
ral
yi
el
d
ge
ner
at
io
n.
T
he
en
ha
nced
yi
el
d
of
the
rice
c
rop
de
pends
la
rg
el
y
on
the
w
at
er
avail
a
bili
t
y
an
d
cl
i
m
atic
con
dit
ion
s
.
F
or
e
xa
m
ple,
low
pre
ci
pitat
ion
or
t
e
m
per
at
ure
ext
rem
es
can
dr
a
sti
cal
ly
di
m
inish
rice
yi
el
d.
Grow
i
ng
bette
r
strat
e
gies
to
foresee
yi
el
d
eff
ic
ie
ncy
in
a
m
ixture
of
cl
i
m
at
ic
c
onditi
ons
can
help
to
unde
rstan
d
the
ro
le
of
dif
fer
e
nt
pri
nciple
fac
tors
t
hat
infl
ue
nce
the
rice
cr
op
yi
el
d.
Bi
g
da
ta
analy
ti
c
m
e
thods
relat
ed
to
the
r
ic
e
crop
yi
el
d
pr
e
dicti
on
a
nd
est
i
m
ation
wil
l
certai
nly
su
pport
the
far
m
e
rs
to
unde
rstan
d
th
e
op
ti
m
u
m
co
ndit
ion
of the si
gnific
ant f
act
or
s
for t
he ric
e cr
op
yi
el
d,
h
e
nce
c
an
ac
hieve
h
i
gher
cr
op yi
el
d.
2
.
1.
Crop
Yiel
d Predicti
on Usin
g M
achi
ne Lear
ning
A
resea
rch
group
in
vestigat
ed
the
util
iz
at
ion
of
var
i
ous
inf
or
m
at
ion
m
ining
m
et
ho
ds
wh
ic
h
will
foresee
rice
cr
op
yi
el
d
f
or
th
e
data
colle
ct
ed
f
ro
m
t
he
sta
t
e
of
Ma
har
as
ht
ra,
I
nd
ia
.
A
t
otal
of
27
re
gio
ns
of
Ma
har
as
htra
w
ere
sel
ect
ed
for
the
assessm
e
nt
and
the
dat
a
was
colle
ct
ed
relat
ed
to
th
e
pr
inciple
ric
e
cro
p
yi
el
d
influ
enci
ng
par
am
et
ers
su
ch
as
diff
e
ren
t
at
m
os
ph
e
ric
conditi
ons
and
var
i
ou
s
ha
rv
est
pa
ram
eter
s
i.e
Pr
eci
pitat
ion
r
at
e
,
m
ini
m
u
m
,
aver
a
ge,
m
axi
m
u
m
and
m
os
t
extrem
e
tem
per
at
ur
e,
re
f
eren
ce
trim
culti
vab
le
area,
e
vapotra
nspirat
io
n,
a
nd
yi
el
d
fo
r
t
he
se
aso
n
bet
ween
J
un
e
t
o
N
ovem
ber
ref
e
rr
e
d
as
Kh
a
rif,
f
or
the
ye
ar
s
1998 to 200
2
f
ro
m
the o
pe
n
s
ource, In
dian Ad
m
inist
rati
on r
eco
r
ds.
WEKA
a Java base
d dial
ect
p
r
ogra
m
m
ing
for
le
ss
chall
e
ng
i
ng
assist
an
ce
with
in
form
at
ion
data
set
s
,
assig
ni
ng
des
ign
outc
om
es
too
l
was
ap
plie
d
for
dataset
proce
s
sing
a
nd
the
ov
e
rall
m
et
ho
do
l
og
y
of
the
stud
y
i
nclu
de
s,
(
1)
pre
-
proc
essing
of
dataset
(2)
Bui
ldin
g
the
predict
io
n
m
od
el
util
iz
ing
WEK
A
a
nd
(
3)
An
al
yz
in
g
the
ou
tc
om
es.
Cros
s
validat
io
n
s
tud
y
is
carried
ou
t
to
s
cru
ti
nize
ho
w
a
pr
edict
a
ble
inf
or
m
at
ion
m
i
ning
m
e
tho
d
w
il
l
execu
te
on
an
am
big
uous
dataset
.
Stud
y
a
ppli
ed
10
-
fo
l
d
highe
r
cr
os
s
validat
ion
st
udy
de
sign
to
assess
the
data
s
ubset
s
f
or
scree
ning
a
nd
te
sti
ng
.
I
den
ti
f
ie
d
a
nd
colle
ct
ed
i
nfor
m
at
ion
was
ra
ndom
l
y
distrib
uted
into
10
sect
i
ons
w
her
e
in
on
e
dat
a
sect
ion
was
use
d
f
or
te
sti
ng
wh
il
e
al
l
oth
er
data
sect
ion
s
wer
e
util
iz
ed
fo
r
the
pr
e
parat
ion
i
nfo
rm
at
io
n.
Stu
dy
repor
te
d
th
at
the
m
et
ho
d
ap
pl
ie
d
was
sup
portive
in
the
pre
ci
se
est
i
m
at
ion
of
rice
cr
op
yi
el
d
f
or
t
he
sta
te
of
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Agricultu
re
Data A
na
ly
ti
cs in
Crop Yie
ld
Esti
ma
ti
on:
A Cri
ti
cal Revi
ew
(
B
M Sa
ga
r
)
1089
Ma
har
as
htra,
I
nd
ia
.
The
pr
eci
se
qu
a
ntific
at
ion
of
t
he
rice
pro
duct
ivit
y
in
var
i
ou
s
cl
im
ati
c
conditi
ons
ca
n
hel
p
far
m
er to
unde
rs
ta
nd the
opti
m
u
m
co
nd
it
ion f
or
t
he hig
her
rice cr
op yi
el
d [8
]
.
Agricult
ure
is
on
e
of
t
he
m
ajo
r
re
venue
pro
du
ci
ng
sect
ors
of
I
nd
ia
an
d
a
so
urce
of
s
urvi
val.
Var
i
ous
seaso
nal,
ec
onom
ic
and
bi
olo
gical
factors
i
nf
l
uen
ce
the
c
r
op
pro
duct
io
n
bu
t
unpredict
a
ble
ch
a
nges
in
these
factors
le
ad
t
o
a
gr
eat
l
os
s
to farm
ers.
These r
isks
ca
n
be
m
easur
e
d
w
hen
s
uitable
m
at
he
m
at
ic
al
and
sta
ti
sti
cal
m
od
el
design
s
are
ap
plied
on
data
relat
ed
to
so
il
,
weathe
r
an
d
past
yi
eld
.
W
it
h
the
a
dvent
of
data
m
ining,
crop
yi
el
d
can b
e
pred
ic
te
d
by
der
ivin
g
us
ef
ul
insig
hts
fro
m
these
agr
ic
ul
tural
data
that ai
ds
far
m
ers
to
decide
on
the
c
rop
th
ey
wo
ul
d
li
ke
to
plant
f
or
the
fo
rt
hco
m
ing
ye
ar
le
adin
g
to
m
axi
m
u
m
pr
ofi
t.
Ther
e
are
va
rio
us
syst
e
m
s
that
use
di
ver
se
da
ta
m
ining
te
ch
nolog
ie
s
t
o
m
anip
ulate
data
to
der
i
ve
in
sig
hts
an
d
help
in
de
ci
sio
n
m
aking
f
or
fa
r
m
ers.
T
he
pr
es
ent
data
m
ining
syst
em
s
and
al
gorithm
s
us
ed
wer
e
f
ocus
e
it
her
on
one
c
r
op
an
d
pr
e
dict
or
f
or
e
cast
any
one
pa
ram
et
er
li
ke
ei
ther
yi
el
d
or
pri
ce.
A
resea
rc
h
prese
nts
a
s
urvey
on
t
he
va
rio
us
al
gorithm
s
us
ed
f
or
cr
op
yi
el
d
predict
io
n,
st
ud
y
use
d
to
forecast
the
yi
eld
an
d
pri
ce
of
m
ajo
r
c
rops
of
Ta
m
il
Nadu
base
d
on
histor
ic
al
da
ta
.
The
data
a
nd
pre
dicte
d
outp
ut
are
acce
ssible
f
or
the
f
arm
ers
throu
gh
a
we
b
app
li
cat
io
n.
T
his
ai
ds
far
m
er
to
decide
on
the
cro
p
they
would
li
ke
to
plant
for
the
forthc
om
ing
year
.
I
n
add
it
io
n,
t
he
w
eb
ap
plica
ti
on
al
so
pro
vi
des
a
foru
m
for
the
far
m
ers
to
go
ods
the
pro
duct
s
without
m
idd
lem
en
wh
ic
h help t
he
m
to obtai
n
m
axim
u
m
p
rice f
or their
pro
duc
ts
[9
]
.
2
.
2
.
Crop
Yiel
d Predicti
on Usin
g Dat
a Mini
ng Te
chn
iques
India
is
a
co
un
t
ry
wh
e
re
f
arm
ing
and
a
gr
ic
ultur
e
bas
ed
in
du
st
ries
are
the
m
ajo
r
resour
ce
of
econom
y.
It
is
al
so
on
e
of
t
he
co
untry
w
hic
h
suffe
r
from
m
ajor
natu
ral
cal
a
m
ities
li
ke
dr
ought
or
flo
od
wh
ic
h
da
m
ages
the
crop
wh
ic
h
cau
se
huge
fina
nc
ia
l
loss
for
th
e
far
m
ers
an
d
econom
ic
stab
il
it
y
of
the
c
ountry
.
Pr
e
dicti
ng
the
crop
yi
el
d
well
in
adv
ance
pr
io
r
to
it
s
harvest
can
he
lp
the
far
m
e
rs
an
d
G
ov
e
r
nm
ent
orga
nizat
ion
s
to
m
ake
a
ppr
opriat
e
plann
i
ng
li
ke
st
or
i
ng
,
sel
li
ng
,
fixing
m
ini
m
u
m
su
pport
pr
ic
e
,
i
m
po
rtin
g/ex
portin
g
et
c.
P
re
dicti
ng
a
cr
op
well
in
ad
van
c
e
requires
a
sy
stem
at
ic
s
tud
y
of
huge
data
com
ing
from
var
iou
s
va
riables
li
ke
s
oil
qu
al
it
y,
pH
,
essenti
al
el
em
ents
(N
,
P,
K)
qu
a
ntit
y
et
c.
As
P
red
ic
ti
on
of
c
rop
deals
with
la
r
ge
set
of
databa
se
thu
s
m
aking
this
pr
edict
io
n
syst
e
m
a
per
fe
ct
cand
idate
f
or
ap
plica
ti
on
of
data
m
ining
m
et
ho
do
l
og
ie
s
w
hich
m
ajo
rly
hel
ps
in
acq
uiri
ng
a
knowle
dg
e
to
achie
ve
hi
gh
e
r
cr
op
yi
el
d.
T
he
su
ccess
of
a
ny
crop
yi
el
d
pr
e
di
ct
ion
syst
em
heav
il
y
reli
es
on
how
acc
urat
el
y
the
featu
res
hav
e
bee
n
e
xtr
act
ed
and
how
a
ppr
opriat
el
y
cl
assifi
ers
ha
ve
bee
n
em
plo
ye
d.
S
tud
y
s
umm
arizes
the
res
ults
ob
ta
ine
d
by
va
rio
us
al
gorithm
s
wh
ic
h
are
bein
g
us
e
d
by
va
rio
us
aut
hors
for
crop
yi
el
d
pr
edict
ion
,
with
the
ir
accu
rac
y
and
reco
m
m
end
at
ion
[
10]
.
Weeds
an
d
pe
sts
wer
e
the
m
ajor
cr
op
dam
agin
g
bi
otic
ag
ents
an
d
the
fa
rm
ers
are
need
to
be
well
-
inf
or
m
ed
in
acce
ssing
t
he
va
rio
us
data
m
i
ning
te
ch
no
l
ogie
s
to
acq
uir
e
a
know
le
dg
e
on
a
ppli
cat
i
on
s
of
eff
ect
ive
wee
d and pes
t c
ontr
ol
strategie
s and
m
anag
ing
tec
hniq
ues
to
r
e
duc
e crop dam
age.
Coll
ect
ion
of
data
relat
ed
to
the
va
rio
us
wee
ds
a
nd
pest,
m
od
el
ing
of
the
data
to
pr
e
par
e
for
the
m
ining
,
sel
e
ct
ion
of
ap
pro
pr
ia
te
m
et
ho
dolo
gy,
interp
retat
ion
and
s
har
i
ng
th
e
inform
at
ion
b
ecom
e
the
m
ajor
chall
en
ge
s
in
weed
a
nd
pest
con
t
ro
l
t
o
prot
ect
the
c
rop
da
m
age.
A
stu
dy
was
co
nducte
d
to
e
valuate
t
he
m
ajo
r
c
halle
ng
e
s
a
nd
note
worthy
opport
un
it
ie
s
and
a
ppli
cat
ion
s
of
of
Bi
g
Dat
a
in
con
t
ro
ll
in
g
the
wee
d
a
nd
pest
da
m
age
and
he
nce
to
a
chieve
higher
c
r
op
yi
e
ld.
St
ud
y
report
ed
that
the
f
orm
of
the
data
c
ollec
te
d,
ty
pe
of
t
he
asses
sm
ent
m
et
ho
d
a
nd
to
ols
app
li
ed
a
re
the
m
ajo
r
in
flue
nc
ing
fact
or
s
i
n
unde
rstan
ding
t
he
r
ole
of
cr
op
dam
aging
age
nts
su
c
h
as
we
ed
a
nd
pest,
wh
ic
h
pro
vid
es
the
knowle
dge
on
us
in
g
im
pr
ove
d
cr
op
m
anage
m
ent
strat
egi
es
an
d
c
rop
yi
el
d
pr
e
dicti
on.
Bi
g
Data
carg
o
s
pa
ce
and
quest
ion
i
ng
inc
urs
intense
ch
al
le
nges,
in
res
pect
to
al
locat
e
th
e
data
acro
s
s
num
ero
us
te
ch
nolo
gies,
an
d
al
so
co
nt
inu
ously
ev
ol
ving
data
from
div
e
rse
s
ource
s.
Wh
e
n
the
se
le
ct
ed
data
was
from
the
diff
e
ren
t
so
urces
,
sem
a
ntic
m
et
ho
do
l
og
ie
s
play
a
vi
ta
l
ro
le
in
the
assessm
ent,
wh
ic
h
pr
el
im
inaril
y
detect
the
factor
s
posses
s
po
te
ntial
agr
ic
ultur
al
im
po
rtance
an
d
de
ve
lop
in
g
relat
io
ns
hi
ps
betwee
n
data
it
e
m
s
in
te
r
m
s
of
m
eanings
a
nd
unit
s.
Stu
dy
pr
ese
nte
d
a
s
uc
cess
story
fro
m
the
Nethe
rland
s
in
us
in
g
t
he
in
for
m
at
ion
from
the
Bi
g
Data
ana
ly
ti
cs,
with
nu
m
erical
al
go
rit
hm
s
in
co
ntro
ll
ing
t
he
c
rop
da
m
age
and
re
porte
d
the
highe
r
crop
yi
el
d.
Stud
y
con
cl
ud
e
d
that
,
the
util
it
y
an
d
the
ap
plica
tio
ns
a
nd
of
Bi
g
data
analy
ti
cs f
or we
ed
a
nd
pest c
ontr
ol is v
e
ry lar
ge
a
nd
pa
rtic
ularly
f
or in
vasi
ve
, p
a
rasit
ic
and
h
er
bicide
-
resis
ta
nt
weeds.
Als
o
im
po
rted
the
ne
ed
of
colla
bora
ti
on
of
ag
ricult
ur
al
sci
entist
s
with
data
sci
en
ti
sts
to
i
m
ple
men
t
t
he
m
et
ho
dolo
gies
for
the
b
e
ne
fit of ag
ricult
ural
pr
act
ic
es
[6]
.
Data
m
ining
play
s
a
piv
ot
al
ro
le
fo
r
de
ci
sion
m
aking
on
dif
fer
e
nt
con
ce
r
ns
with
resp
ect
to
agr
ic
ultur
e
pra
ct
ic
es.
The
obje
ct
ive
of
the
da
ta
m
ining
m
eth
ods
is
t
o
m
ine
knowle
dge
fro
m
an
acce
s
sible
data
set
and
c
onve
r
t
it
into
a
com
pr
e
he
ns
ible
f
or
m
at
fo
r
s
om
e
sign
ific
a
nt
ap
pl
ic
at
ion
of
t
he
Agri
proce
ss.
Crop
m
anag
em
ent
of
certai
n
ag
ric
ultur
e
reg
i
on
i
s
dep
e
ndin
g
on
the
cl
i
m
at
ic
conditi
on
s
of
that
reg
io
n
be
cause
cl
i
m
at
e
can
m
ake
huge
im
pact
on
cr
op
pr
oductivit
y.
Re
al
tim
e
weather
data
can
help
to
achieve
the
good
crop
m
anag
e
m
ent.
Ef
fecti
ve
util
iz
at
ion
of
m
ined
ag
ricult
ur
al
base
d
inf
orm
ation
and
c
omm
un
i
cat
ions
exp
e
rtise
ena
bl
es
autom
ation
of
retrie
ving
us
ef
ul
data
in
an
ef
fort
to
acqu
i
re
knowle
dge,
w
hic
h
pro
vid
es
opport
un
it
y
to
easi
er
data
acq
uisit
ion
f
ro
m
e
le
ct
ro
nic
sou
rc
es
directl
y,
transf
e
r
to
secu
re
el
ect
ro
nic
syst
e
m
of
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
12
, N
o.
3
,
Dece
m
ber
2
01
8
:
1
0
87
–
1
0
93
1090
do
c
um
entat
ion
and
re
du
ce
s
m
anu
al
ta
sk
s.
Au
t
om
ation
strat
egies
re
du
c
e
the
ov
er
al
l
pr
od
uction
c
os
t,
hen
ce
su
pp
or
t
f
or
hig
he
r
cr
op
yi
el
d
an
d
highe
r
m
ark
et
pr
ic
e.
Also
ide
ntifie
d
that
ho
w
the
data
m
ining
he
lps
t
o
analy
ze
and
predict
the
us
ef
ul
patte
rn
f
rom
hu
ge
an
d
dy
nam
ic
al
ly
chan
ge
d
cl
i
m
at
ic
data.
In
the
f
ie
ld
of
agr
ic
ultur
al
bi
oenginee
rin
g,
sci
entist
an
d
eng
i
neer
s
in
c
ollaborat
io
n
ha
ve
devel
oped
an
d
discusse
d
the
app
li
cat
io
n
of
m
at
he
m
at
ic
a
l m
od
el
d
esi
gns
li
ke
f
uzzy lo
gic d
esi
gns
in opt
i
m
iz
at
ion
of t
he
cr
op yi
el
d,
arti
fici
al
neural
net
wor
ks
in
validat
io
n
stu
dies,
genet
ic
al
go
rithm
s
desig
ns
in
a
ccessi
ng
t
he
f
it
ness
of
t
he
m
od
el
app
li
e
d,
decisi
on
trees,
as
we
ll
as
suppo
rt
ve
ct
or
m
achines
to
asse
ss
s
oil,
cl
i
m
at
e
con
dit
ion
s
an
d
a
vaila
bili
ty
of
water
re
sources
relat
ed
t
o
cr
op
gro
wth
and
pest
m
a
nag
em
ent
in
agr
ic
ultur
e.
Stu
dy
su
m
m
arize
s
the
app
li
cat
io
n
of
d
at
a
m
ining
te
chnolo
gies
i.e
Neural
Netw
orks,
S
upport
Ve
ct
or
Ma
chine
,
Bi
g
Data
analy
sis
an
d
so
ft c
om
pu
ti
ng in
the
assessm
ent of a
gr
ic
ultu
re f
ie
ld
b
a
sed
on weat
he
r
c
ondi
ti
on
s
[
5].
2
.
3
.
Crop
yield predi
cti
on
using
Bi
g Dat
a An
aly
tics
In
I
ndia
crop
yi
el
d
is
season
de
pende
nt
and
m
ajo
rly
inf
luence
d
by
th
e
bio
lo
gical
and
ec
onom
ic
causes
of
a
n
i
nd
i
vidual
cr
op.
Re
portin
g
of
pro
gr
essi
ve
a
gri
cultural
yi
el
d
in
al
l
the
seas
on
s
is
an
am
ple
ta
sk
and
a
n
a
dv
a
nt
ageous
ta
sk
f
or
e
ver
y
natio
n
with
res
pe
ct
t
o
asse
sses
t
he
overall
c
rop
yi
el
d
pr
e
dicti
on
a
nd
est
i
m
ation
.
At
pr
ese
nt
a
com
m
on
issue
wo
r
ldwid
e
is,
fa
rm
ers
are
stresse
d
in
pr
od
ucin
g
higher
cr
op
yi
el
d
du
e
to
the
in
flue
nc
e
of
unpredict
a
ble
cl
i
m
at
ic
ch
ang
e
s
an
d
sig
ni
ficant
reducti
on
of
water
re
so
urce
world
w
ide.
A
stud
y
was
car
r
ie
d
ou
t
to
coll
ect
the
data
on
wo
rl
d
cl
i
m
at
i
c
changes
an
d
the
avail
able
water
res
ource
s
wh
ic
h
can
be
us
e
d
t
o
enc
oura
ge
adv
a
nce
d
a
nd
novel
ap
proa
ches
s
uch
a
s
big
data
anal
yt
ic
s
to
retrieve
the
inf
or
m
at
ion
of
the
previ
ou
s
resu
lt
s
t
o
t
he
crop
yi
el
d
pre
dicti
on
a
nd
e
s
tim
a
ti
on
.
Stu
dy
i
m
po
rted
th
at
the
sel
ect
ion
an
d
us
a
ge
of
the
m
os
t
desirab
le
cro
p
acc
ordin
g
to
the
existi
ng
c
onditi
ons,
su
pp
or
t
to
ach
ie
ve
the
higher
and e
nh
anced cr
op yi
el
d
[11].
The
acc
ur
at
e
predict
io
n
of
c
r
op
yi
el
d
ce
rta
inly
be
nef
it
s
th
e
far
m
ers
in
c
hoos
i
ng
the
rig
ht
m
et
ho
d
to
reduce
t
he
c
rop
dam
age
and
gets
best
pr
ic
es
f
or
their
c
r
op
s
.
A
researc
h
gro
up
co
nduc
te
d
a
wor
k
w
it
h
a
n
obj
ect
ive
of
a
ccur
at
e
pr
e
dic
ti
on
of
c
r
op
yi
el
d
thr
ough
big
data
a
na
ly
ti
cs
to
assess
va
rio
us
c
ro
p
yi
el
d
influ
e
ncin
g
fa
ct
or
s
s
uch
as
Ar
ea
under
C
ul
ti
vation
(AU
C)
interi
m
s
of
hector
s
,
A
nnua
l
Ra
infall
(AR
)
rates
and
F
ood
P
ric
e
Index
(
FP
I)
and
to
de
velo
p
relat
ion
s
hip
a
m
on
g
these
pa
ram
et
ers.
Re
gr
ession
An
al
ysi
s
(RA)
m
et
ho
dolo
gy
was
a
pp
li
e
d
to
exam
ine
the
se
le
ct
ed
facto
rs
a
nd
their
im
pact
on
cr
op
pr
e
dic
ti
on
a
nd
fi
nal
yi
el
d.
RA
m
e
tho
dol
ogy
is
a
m
ulti
var
ia
ble
in
vestig
at
ion
pract
ic
e
wh
ic
h
can
cat
e
gorize
the
fact
or
s
i
n
to
gr
oups
su
c
h
as
exp
la
nat
or
y
and
res
pons
e
var
ia
bles
an
d
helps
to
assess
their
interac
ti
on
to
ob
ta
i
n
a
reso
luti
on.
All
the
sel
ect
ed
factors
of
the
prese
nt
stud
y
desig
n
known
as
AR
,
AU
C
an
d
FP
I
wer
e
m
easur
ed
for
a
per
i
od
of
10
ye
ars
bet
ween
the
ye
ars
1990
-
20
00.
A
nove
l
m
e
tho
d
cal
le
d
Linear
Re
gr
e
s
sion
(LR)
is
ap
plied
to
a
naly
z
e
the
relat
ion
s
hip
be
tw
een
e
xp
la
na
tory
var
ia
ble
s
(A
R,
AU
C,
FPI
)
a
nd
t
he
crop
yi
el
d
con
si
der
e
d
as
r
esp
on
se
var
ia
ble.
St
ud
y
re
ported
that
t
he
R
2
val
ue
for
the
st
ud
ie
d
fa
ct
or
s
cl
ea
rly
in
dicat
e
that
c
rop
yi
el
d
is
pri
nc
ipall
y
dep
e
nds
on
A
R.
Stu
dy
al
so
repor
te
d
that
t
he
oth
er
tw
o
f
act
or
s
(AUC
a
nd
FP
I)
sc
ree
ne
d
wer
e
al
s
o
f
ound
t
o
hav
e
sig
nifican
t
i
m
pact
after
the
AR
.
Stu
dy
sh
al
l
be
c
on
ti
nued
t
o
analy
ze
the
i
m
pact
of
for
ot
her
s
ubst
antia
l
factors
li
ke
Mi
nim
u
m
Su
pport
Pr
ic
e
(MSP
),
Cost
Pr
ic
e
Index
(CP
I)
,
Wholesal
e
Pr
ic
e
I
nd
e
x
(
WPI)
et
c.
a
nd
their r
el
at
io
nsh
ip on
t
he
yi
el
ds o
f diffe
ren
t c
r
op
s
[1
2].
Crop
yi
el
d
ga
ps,
m
easur
ed
as
dif
fer
e
nce
between
ex
pected
yi
el
ds
base
d
on
the
po
te
ncy
an
d
act
ua
l
far
m
yi
el
d
received.
I
n
or
der
to
achie
ve
the
higher
c
r
op
y
ie
ld,
fa
rm
ers
m
us
t
need
to
ta
ckle
the
in
flu
encin
g
factors
su
c
h
as
influ
e
nce
of
change
in
cl
im
a
te
con
diti
ons
on
the
pros
pects
of
cr
op
yi
el
ds,
and
c
hange
i
n
the
us
a
ge
of
a
gr
ic
ultur
al
la
nd
t
o
assess
a
nd
ult
i
m
at
ely
red
uce
the
cr
op
yi
el
d
ga
ps
.
Se
ve
ral
researc
hers
re
porte
d
the
ap
plica
ti
ons
of
bio
sim
u
la
ti
on
m
od
el
s
to
est
i
m
at
e
the
cro
p
yi
el
d
ga
ps
in
the
la
st
deca
de.
The
im
pact
of
th
e
crop
yi
el
d
ga
ps
assessm
ent
s
tud
ie
s
co
nduct
ed
thr
ough
bio
si
m
ulati
on
ba
sed
m
et
ho
do
l
ogie
s
we
re
ne
ga
ti
vely
influ
e
nce
d
by
qu
al
it
y
and
res
olu
ti
on
of
cl
i
m
at
e
and
s
oil
data,
as
we
ll
as
un
s
ci
entifi
cal
ly
exp
ect
at
ions
about
crop
yi
el
d
pre
dicti
on
syst
em
s
an
d
cr
op
yi
el
d
assessm
ent
m
od
el
ing
desi
gn
s
cal
ibrati
on
m
et
ho
d.
A
n
e
xp
li
ci
t
rati
on
al
e
m
od
e
l
wh
ic
h
can
ef
fecti
vely
ap
plied
at
va
rio
us
l
evels
of
the
av
ai
la
bili
ty
of
qual
it
y
info
rm
at
i
on
f
o
r
identify
in
g
data
sources
to
a
na
ly
ze
crop
yi
el
d
a
nd
m
easur
in
g
yi
el
d
gaps
at
d
efi
nite
ge
ogra
ph
ic
al
loc
at
ions
an
d
works
base
d
on
the
rise
in
ti
t
er
appr
oach.
The
m
od
el
is
hi
gh
ly
help
fu
l
in
retrievin
g
the
us
ef
ul
data
fro
m
the
avail
able,
poor
qual
it
y,
le
ss
rig
or
ou
s
data
s
ources
or
if
t
he
data
is
not
a
va
il
able.
A
cas
e
stud
y
was
disc
us
se
d
on
the
a
pp
li
cat
ion
of
sel
ect
ed
m
od
el
desig
n
to
qu
a
ntify
the
yi
el
d
gap
s
of
m
ai
ze
cro
p
in
the
sta
te
of
Ne
br
as
ka
(U
S
A
),
an
d
al
so
at
the
dif
fe
ren
t
ge
ogra
phic
al
locat
ion
s
r
epr
ese
ntin
g
th
e
nations
Arg
entina
an
d
Ke
nya
at
national
scal
e
le
vel.
Di
ff
e
rent
geog
raphical
locat
ions
s
uc
h
as
Ne
br
as
ka
(USA
),
A
rg
e
ntina
a
nd
Ke
ny
a
were
identifie
d
t
o
s
ym
bo
li
ze
the
di
sti
nct
scenarios
of
A
gr
i
base
d
data
a
vaila
bi
li
ty
and
the
qu
al
it
y
fo
r
the
se
le
ct
ed
var
ia
bles
asses
sed
t
o
pr
e
dict
an
d
est
im
at
e
the
c
rop
yi
el
d
gaps.
The
de
finiti
ve
as
pir
at
ion
of
the
pl
anne
d
m
et
ho
d
is
to
afford
tran
sp
a
re
nt,
easi
ly
accessib
le
,
re
produ
ci
ble
and
te
ch
nical
ly
so
und
and
str
ong
gu
i
delines
for
predict
in
g
the
yi
el
d
gap
s
.
The
pr
opos
e
d
g
ui
delines
we
re
al
so
rele
vant
fo
r
unde
rstan
ding
an
d
to
si
m
ula
te
the
infl
uen
ce
of
cha
nge
in
cl
im
at
e
con
diti
on
s
an
d
us
a
ge
of
culti
vab
le
la
nd
cha
nges
f
r
om
national
to
global
scal
es.
As
in
di
cat
ed,
the
bette
r
unde
rsta
nd
i
ng
of
data
im
portance
a
nd
us
e
f
uln
ess
for
a
naly
zi
ng
cr
op
yi
el
d
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Agricultu
re
Data A
na
ly
ti
cs in
Crop Yie
ld
Esti
ma
ti
on:
A Cri
ti
cal Revi
ew
(
B
M Sa
ga
r
)
1091
est
i
m
ating
yi
eld
ga
ps
as
il
lustrate
d
can
he
lp
in
identify
in
g
the
data
ga
ps
in
the
cro
p
yi
el
d
and
al
low
f
oc
us
in
g
on the
va
rio
us
effor
ts t
a
ken at
the
global l
eve
l t
o
ad
dress t
he
m
os
t crit
ic
al
is
su
e
[13].
An
al
yz
in
g
the
yi
el
ds
of
crop
i
s
necessary
to
update
the
po
l
ic
ie
s
to
ensu
re
fo
od
sec
ur
it
y.
A
researc
h
gro
up
c
onduct
ed
a
stud
y
wit
h
the
ai
m
in
su
gg
e
sti
ng
a
no
ve
l
data
m
ining
m
et
ho
d
to
pr
e
di
ct
the
yi
e
lds
of
crop
dep
e
nds
on
a
gr
ic
ultur
al
bi
g
data
analy
t
ic
s
m
e
tho
dolo
gies,
w
hich
wer
e
pro
gress
ively
con
t
rast
with
conve
ntion
al
data
m
ining
m
et
ho
dolo
gies
in
the
proc
ess
of
ha
ndli
ng
data
a
nd
m
od
el
in
g
de
signs
.
Stud
y
sug
gest
ed
that
the
m
et
hod
em
plo
yed
shou
l
d
be
use
r
fr
ie
nd
ly
,
work
base
d
on
pro
gr
es
sive
big
-
data
respo
ns
ive
pro
cessi
ng
struct
ure,
s
uppose
d
to
util
iz
e
the
e
xisti
ng
a
gr
ic
ul
tural
base
d
si
gnific
ant
datase
ts
and
would
sti
ll
be
us
e
d
with
the
l
arg
e
r
vo
l
um
es
of
d
at
a g
r
ow
i
ng
at
en
orm
ou
s r
at
es. N
earest neig
hbors
m
odel
ing
is
on
e
s
uch
nove
l
data
m
ining
te
chn
i
qu
e
w
hich
w
orks
on
the
res
ults
colle
ct
ed
base
d
on
data
pro
cessi
ng
structu
res
f
or
m
the
far
m
ers
and
sugg
e
st
a
well
un
biase
d
res
ult
on
the
base
of
accu
racy
and
pr
e
dicti
on
t
i
m
e
in
adv
a
nce.
Stu
dy
furthe
r
discu
ssed
a
case
st
udy
on
t
he
asse
ssm
ent
of
act
ua
l
crop
dataset
(
nu
m
erical
ex
a
m
ples
on)
in
C
hin
a
f
ro
m
19
95
-
20
14.
St
ud
y
re
por
te
d
that
the
no
vel
m
od
el
e
m
plo
ye
d
has
publici
zed
an
im
pro
ve
d
perform
ance
and
was
f
ound
to
be
pro
gressi
ve
in
re
portin
g
pr
e
dicti
on
acc
ur
acy
per
ce
nta
ge
of
t
he
c
ompare
d
m
et
ho
dolo
gies
with c
onve
ntio
nal d
e
sig
ns
[7
]
.
Si
m
ulati
on
m
od
el
s
base
d
on
fiel
d
ex
pe
rim
ent
are
valuab
le
te
ch
nolo
gies
f
or
st
ud
yi
ng
a
nd
unde
rstan
ding
crop
yi
el
d
gaps,
but
one
of
t
he
crit
ic
al
chal
le
ng
e
rem
ai
n
with
these
m
et
hods
is
scal
in
g
up
of
these
ap
proac
h
to
assess
the
data
colla
te
d
be
tween
di
ff
e
re
nt
tim
e
interva
ls
fr
om
the
broad
e
r
ge
ogra
phic
al
reg
i
on
s
.
Sate
ll
it
e
retrieved
da
ta
hav
e
f
re
quently
been
re
ve
al
ed
to
prese
nt
data
set
s
that,
by
it
sel
f
or
i
n
gro
up
i
ng
with
oth
e
r
inf
or
m
at
i
on
a
nd
m
od
el
desig
ns
,
ca
n
pr
eci
sel
y
deter
m
i
ne
the
yi
el
ds
of
crop
in
a
gr
ic
ultur
a
l
la
nd
s
.
The
yi
el
d
m
aps
dev
el
oped
s
hall
pro
vi
de
an
un
i
que
oppo
rtu
nity
to
ov
erc
om
e
bo
th
sp
at
ia
l
and
te
m
po
ral
base
d
scal
in
g
up
chall
en
ges
and
th
us
im
pr
ov
e
the
i
deo
l
ogy
of
cr
op
yi
el
d
ga
ps
pr
e
dic
ti
on
.
A
rev
ie
w
was
cond
ucted
to
di
scuss
t
he
a
pp
l
ic
at
ion
s
of
re
m
ote
sensing
te
chnolo
gy
t
o
determ
ine
the
i
m
pact
and
ca
us
es
of
yi
el
d
gap
s.
E
ve
n
th
ough
the
exam
ple
discu
ssed
by
the
res
earch
gro
up
de
m
on
strat
es
the
us
ef
uln
es
s
of
rem
ote
sensing
in
the
pr
e
dicti
on
of
y
ie
ld
gap
s
,
but
al
so
m
any
are
as
of
possi
ble
app
li
cat
io
n
with
res
pect
to
th
e
crop
y
ie
ld
assessm
e
nt,
pre
dicti
on
and
im
pr
ovem
ent
rem
ai
n
unexp
l
or
e
d.
Stu
dy
pr
op
os
e
d
tw
o
le
ss
com
plica
te
d,
easi
ly
assessable
m
et
ho
ds
t
o
determ
ine
and
qua
ntify
th
e
yi
el
d
ga
ps
betwee
n
var
i
ous
a
gr
ic
ultur
al
fiel
ds
.
First
m
et
ho
d
works
cl
os
el
y
with
the
const
r
uctive
m
a
ps
rep
rese
ntin
g
the
aver
a
ge
cr
op
y
ie
lds,
it
can
be
us
ed
directl
y
to
acc
esses
s
pecific
crop
yi
el
d
i
nf
l
uen
ci
ng
fact
ors
f
or
f
ur
t
her
stud
ie
s
w
her
ea
s
the
sec
ond
m
eth
od
use
the
rem
ote
sensing
te
ch
nolo
gy
to
retrieve
the
data
f
or
prov
i
ding
the
us
efu
l
in
form
at
i
on
r
ega
r
ding
the
cr
op
yi
el
d
pr
e
dicti
on a
nd esti
m
a
ti
on
[14].
In
com
ing
dec
ades,
tw
o
m
os
t
sign
ific
a
nt
a
nd
im
po
rtant
f
act
or
s
f
ound
t
o
in
flue
nce
c
r
op
yi
el
d
is,
increase
i
n
th
e
global
popula
ti
on
a
nd
e
conom
y,
wh
ic
h
gr
eat
ly
de
m
and
s
t
he
hig
he
r
a
nd
s
us
ta
inabl
e
agr
ic
ul
tural
ba
sed
cr
op
yi
el
ds.
The
capaci
ti
es
of
f
ood
pro
du
ct
io
n
at
global
le
vel
is
goi
ng
t
o
be
ve
ry
lim
i
te
d
du
e
to
the
le
ss
avail
abili
ty
of
culti
vab
le
la
nd
,
water
res
ourc
es,
diff
ic
ulti
es
in
m
ai
ntaining
the
su
sta
ina
ble
cro
p
pro
du
ct
io
n
le
ve
ls,
eff
ect
s
of
changes
in
t
he
gl
ob
al
cl
im
at
ic
con
diti
ons
an
d
al
s
o
by
var
io
us
bioph
ysi
cal
par
am
et
ers
w
hich
i
nf
l
uen
ce
the
c
rop
yi
e
ld.
T
he
far
m
ers
need
to
be
ed
ucated
on
the
a
ppli
cat
i
on
of
sci
entifi
cal
ly
pr
ove
n
m
et
ho
ds
to
qua
ntify
th
e
crop
yi
el
d
ca
pacit
ie
s
an
d
sa
m
e
need
t
o
be
inf
or
m
ed
to
hi
gh
e
r
auth
or
it
ie
s
to
m
ai
ntain
transp
are
ncy
in
sha
rin
g
the
act
ual
info
rm
at
ion
,
intern
helps
i
n
m
aking
the
po
li
cy
base
d,
re
searc
h
ori
ente
d,
de
velo
pm
ent
and
inv
est
m
ent
relat
ed
decisi
ons
that
aim
to
i
nf
l
uen
ce
f
utu
r
e
cr
op
yi
el
d.
Crop
producti
on
abili
ti
e
s
and
yi
el
d
gaps
can
be
a
ssessed
a
nd
m
easur
e
d
by
c
om
par
ing
the
possibl
e
yi
el
ds
at
norm
al
conditi
on
s
with
res
pect
t
o
the
c
rop
pro
du
ct
io
n
unde
r,
resp
ect
i
vely
,
irrigate
d
a
nd
ra
in
fe
d
conditi
ons
by
keep
i
ng
t
he
cr
op
yi
el
d
le
vel
s
lim
it
ed
by
t
he
le
ss
avail
abili
ty
of
the
water
as
be
nc
hm
ark
s.
Yiel
d
ga
ps
ca
n
be
de
fine
d
as
the
dif
fer
e
nce
betwee
n
the
e
xpect
ed
c
rop
yi
el
ds
with
res
pe
ct
to
the
act
ual
crop
yi
el
d
an
d
acc
urat
e,
s
patia
ll
y
un
am
big
uo
us
awar
e
ness
an
d
inf
or
m
at
ion
a
bout
th
e
yi
el
d
gaps
is
necess
ary
to
achieve
s
us
t
ai
nab
le
am
plific
at
ion
of
a
gr
ic
ultur
al
yi
el
ds
.
Keep
i
ng
an
ai
m
of
discuss
i
ng
the
im
pact
of
t
he
var
i
ou
s
m
et
hods
pract
ic
ed
in
m
easur
in
g
the
yi
el
d
gap
s
wit
h
a
spotl
igh
t
on
the
local
-
to
-
global
i
m
po
rta
nce
of
ou
tc
om
es,
a
re
search
gr
oup
c
arr
ie
d
out
a
s
ur
vey
on
t
he
var
i
ou
s
m
et
ho
ds
a
pp
li
ed
to
e
stim
at
e
yi
el
d
gaps.
Stud
y
repor
te
d
few
sta
nd
a
rd
op
e
ra
ti
on
m
e
tho
ds,
e
m
plo
ye
d
in
qu
a
ntifyi
ng
th
e
cro
p
yi
el
d
pote
ntial
on
th
e
data
colle
ct
ed
f
ro
m
the
fa
rm
ers
of
wester
n
Ken
ya
,
Ne
br
a
sk
a
(
U
SA
)
an
d
Vict
ori
a
(
Au
st
rali
a).
Stu
dy
rec
omm
en
de
d
for
the
use
of
accurate
a
nd
r
ecent
yi
el
d
data
assessed
th
r
ough
cal
ib
rated
crop
m
od
el
de
sign
s
a
nd
f
ur
th
er
up
scal
ing
validat
ed
m
et
ho
ds
in
the
predict
io
n
of
cr
op
yi
el
d
ga
ps
T
he
bott
om
-
up
a
ppli
cat
ion
of
this
global
protoc
ol all
ow
s v
e
rificat
ion o
f
est
im
a
te
d
yi
el
d
ga
ps wit
h o
n
-
fa
rm
d
at
a and
exp
e
rim
ents [
15]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
12
, N
o.
3
,
Dece
m
ber
2
01
8
:
1
0
87
–
1
0
93
1092
Table
1.
Dif
fere
nt d
at
a m
inin
g
te
ch
niques
a
pp
li
ed
in
c
r
op
yi
el
d
pr
e
dicti
on
SL
T
i
t
l
e of
t
h
e pa
p
er
A
u
t
h
o
r
s
/
Y
ear o
f
p
u
b
l
i
c
at
i
o
n
H
i
g
h
l
i
g
h
t
s
Ci
t
at
i
o
n
1.
Ri
ce Cr
o
p
Y
i
e
l
d
Pr
ed
i
ct
i
o
n
u
s
i
n
g
D
a
t
a Mi
n
i
n
g
T
ech
n
i
q
u
es
:
A
n
O
v
er
v
i
ew
D
ak
s
h
ay
i
n
i
Pat
i
l
,
D
r.
M .
S,
Sh
i
rd
h
o
n
k
a
r/
2
0
1
7
D
i
s
c
u
s
s
e
d
v
ar
i
o
u
s
d
at
a
m
i
n
i
n
g
t
ec
h
n
i
q
u
es
u
t
i
l
i
z
ed
f
o
r
p
re
d
i
ct
i
o
n
o
f
ri
ce cr
o
p
y
i
el
d
f
o
r
t
h
e
st
a
t
e
o
f Ma
h
ara
s
h
t
r
a,
In
d
i
a.
W
E
K
A
to
o
l
w
a
s
a
p
p
l
i
ed
i
n
d
a
t
a
s
e
t
pr
o
ce
s
s
i
n
g
[8
]
2.
A
Surv
ey
on
Cr
o
p
Y
i
el
d
Pred
i
c
t
i
o
n
b
a
s
e
d
o
n
A
g
ri
c
u
l
t
u
ra
l
D
at
a
D
h
i
v
y
a B
H
,
Man
j
u
l
a
R,
Si
v
a Bh
arat
h
i
S,
Mad
h
u
m
at
h
i
R/
2
0
1
7
Pres
e
n
t
ed
a
s
u
r
v
ey
o
n
t
h
e
d
i
ffe
ren
t
al
g
o
r
i
t
h
m
s
ap
p
l
i
ed
i
n
t
h
e
as
s
es
s
m
en
t
an
d
p
re
d
i
ct
i
o
n
of c
ro
p
y
i
el
d
D
i
s
c
u
s
s
e
d
a
b
o
u
t
t
h
e
m
ech
an
i
s
m
o
f
k
n
o
w
l
e
d
g
e
t
h
e
d
i
s
c
o
v
ery
i
n
A
g
ri
c
u
l
t
u
ra
l
d
at
a m
i
n
i
n
g
[9
]
3
A
Study
on
V
ari
o
u
s
D
at
a
Mi
n
i
n
g
T
ec
h
n
i
q
u
e
s
f
o
r
Cro
p
Y
i
e
l
d
Pr
ed
i
c
t
i
o
n
Y
o
g
e
s
h
G
an
d
g
e,
San
d
h
y
a/
2
0
1
7
D
i
s
c
u
s
s
e
d
v
ari
o
u
s
d
a
t
a
m
i
n
i
n
g
t
ec
h
n
i
q
u
e
s
e
m
p
l
o
y
ed
f
o
r
p
r
ed
i
c
t
i
n
g
t
h
e
cro
p
y
i
el
d
a
n
d
s
i
g
n
i
fi
e
s
t
h
e
i
m
p
o
rt
an
ce
o
f
acc
u
ra
t
e
d
at
a
ex
t
ract
i
o
n
m
et
h
o
d
s
of
b
i
g
d
at
a a
n
al
y
t
i
cs
.
[1
0
]
4.
Bi
g
D
a
t
a f
o
r w
ee
d
co
n
t
r
o
l
an
d
cro
p
p
ro
t
ec
t
i
o
n
F K
V
an
E
v
ert
,
S F
o
u
n
t
a
s
,
D
Jakovet
i
c,
V
Crn
o
j
ev
i
c,
I T
rav
l
o
s
&
C
K
e
m
p
en
aar
/
2
0
1
7
Cri
t
i
c
al
l
y
d
i
s
c
u
s
s
ed
ab
o
u
t
t
h
e
c
h
a
l
l
en
g
es
fac
ed
a
n
d
t
h
e
p
ro
f
o
u
n
d
o
p
p
o
rt
u
n
i
t
i
es
l
i
es
i
n
t
h
e Bi
g
D
a
t
a a
n
al
y
t
i
c
s
in
a
g
r
i
cu
l
t
u
r
e:
O
u
t
l
i
n
e
d
Bi
g
D
at
a an
al
y
t
i
cs
m
o
d
e
l
s
w
i
t
h
n
u
m
eri
cal
a
l
g
o
r
i
t
h
m
s
appl
i
ed
Rep
re
s
e
n
t
t
h
e
i
m
p
o
rt
an
ce
o
f
r
efo
rm
i
n
g
t
h
e
m
i
n
ed
d
at
a
i
n
t
h
e
f
o
rm
o
f
u
n
d
er
s
t
an
d
a
b
l
e i
n
f
o
r
m
at
i
o
n
t
o
t
h
e f
arm
ers
.
D
i
s
c
u
s
s
e
d
a
b
o
u
t
v
ar
i
o
u
s
ad
v
a
n
ce
s
,
t
o
o
l
s
an
d
al
g
o
ri
t
h
m
s
ap
p
l
i
e
d
i
n
t
ra
n
s
f
o
rm
i
n
g
t
h
e
d
at
a
i
n
t
o
ea
s
i
l
y
u
n
d
er
s
t
an
d
a
b
l
e
i
n
f
o
rm
at
i
o
n
t
o
t
h
e
fram
ers
an
d
t
h
ro
w
n
a
l
i
g
h
t
o
n
s
u
cce
s
s
s
t
o
ry
o
f
N
et
h
er
l
a
n
d
s
i
n
ach
i
e
v
i
n
g
t
h
e m
ax
i
m
u
m
crop
y
i
el
d
a
n
d
t
h
e
i
r sm
art
form
i
n
g
p
r
ac
t
i
c
es
.
A
l
s
o
d
i
s
c
u
s
s
e
d
ab
o
u
t
t
h
e
co
n
t
r
o
l
o
f
i
n
v
a
s
i
v
e,
p
ar
as
i
t
i
c
a
n
d
h
er
b
i
c
i
d
e
-
res
i
s
t
a
n
t
w
ee
d
s
t
o
i
m
p
ro
v
e
t
h
e
o
v
era
l
l
c
ro
p
y
i
el
d
ap
p
l
y
i
n
g
B
i
g
D
a
t
a
an
a
l
y
t
i
cs
.
[6
]
5.
T
h
e Im
p
act
of D
a
t
a
A
n
al
y
t
i
c
s
i
n
Cr
o
p
Man
a
g
em
en
t
b
as
ed
o
n
W
eat
h
er C
o
n
d
i
t
i
o
n
s
Sw
aru
p
a
Ra
n
i
A
/
2
0
1
7
D
i
s
c
u
s
s
e
d
t
h
e
a
p
p
l
i
cat
i
o
n
o
f
m
at
h
em
at
i
cal
m
o
d
el
l
i
k
e
fu
z
z
y
l
o
g
i
c
d
es
i
g
n
s
i
n
o
p
t
i
m
i
zat
i
o
n
o
f
t
h
e
c
ro
p
y
i
el
d
,
art
i
fi
ci
a
l
n
e
u
ra
l
n
et
w
o
rk
s
i
n
v
al
i
d
a
t
i
o
n
s
t
u
d
i
e
s
,
g
e
n
et
i
c
al
g
o
ri
t
h
m
s
d
es
i
g
n
s
i
n
acce
s
s
i
n
g
t
h
e
fi
t
n
e
s
s
o
f
t
h
e
m
o
d
e
l
a
p
p
l
i
ed
,
d
ec
i
s
i
o
n
t
ree
s
,
a
n
d
s
u
p
p
o
r
t
v
ec
t
o
r
m
ach
i
n
es
t
o
s
t
u
d
y
s
o
i
l
,
cl
i
m
at
e
co
n
d
i
t
i
o
n
s
a
n
d
w
at
er
re
g
i
m
es
rel
a
t
e
d
t
o
cr
o
p
g
ro
w
t
h
an
d
pe
s
t
m
an
ag
em
en
t
in
a
g
r
i
cu
l
t
u
r
e.
[5
]
6.
A
Study
on
Cro
p
Y
i
e
l
d
Fo
rec
as
t
i
n
g
U
s
i
n
g
Cl
a
s
s
i
f
i
ca
t
i
o
n
T
ec
h
n
i
q
u
es
R.Suj
at
h
a,
D
r.P.
Is
a
k
k
i
D
ev
i
/
2
0
1
6
D
i
s
c
u
s
s
t
h
e
i
m
p
o
rt
a
n
ce
o
f
c
o
m
p
ari
n
g
p
re
v
i
o
u
s
ag
ri
c
u
l
t
u
ra
l
d
a
t
a
w
i
t
h
p
re
s
e
n
t
t
o
i
d
en
t
i
fy
op
t
i
m
u
m
con
d
i
t
i
o
n
fav
o
r en
h
an
ce
d
cro
p
y
i
el
d
.
E
n
v
i
s
a
g
e
d
t
h
e
i
m
p
o
rt
a
n
ce
o
f
b
es
t
cr
o
p
s
el
ect
i
o
n
d
e
p
e
n
d
i
n
g
o
n
t
h
e
s
ea
s
o
n
a
n
d
t
h
e cl
i
m
at
i
c fac
t
o
rs
w
h
i
c
h
s
u
p
p
o
rt
s
e
n
h
an
ced
c
r
o
p
y
i
el
d
.
[1
1
]
7.
Pred
i
c
t
i
o
n
o
f Cr
o
p
Y
i
el
d
u
s
i
n
g
Re
g
re
s
s
i
o
n
A
n
al
y
s
i
s
V
.
Sel
l
am
an
d
E
.
Po
o
v
am
m
al
/
201
6
Reg
re
s
s
i
o
n
an
a
l
y
s
i
s
w
as
carr
i
e
d
o
u
t
t
o
f
i
n
d
t
h
e
re
l
at
i
o
n
s
h
i
p
am
o
n
g
t
h
e
p
aram
et
ers
i
.e
A
rea
u
n
d
er
C
u
l
t
i
v
at
i
o
n
(A
U
C),
A
n
n
u
al
Ra
i
n
fal
l
(A
R)
an
d
Fo
o
d
Pr
i
ce
I
n
d
e
x
(FPI)
w
h
i
c
h
i
n
fl
u
e
n
ce
s
t
h
e
fi
n
a
l
cr
o
p
y
i
el
d
a
n
d
rep
o
rt
ed
t
h
at
t
h
e
cr
o
p
y
i
el
d
p
r
i
n
c
i
p
a
l
l
y
d
ep
en
d
s
o
n
t
h
e
A
n
n
u
al
R
ai
n
fal
l
(A
R).
[1
2
]
8
H
o
w
goo
d
i
s
g
o
o
d
e
n
o
u
g
h
?
D
at
a re
q
u
i
rem
en
t
s
fo
r
rel
i
ab
l
e cr
o
p
y
i
el
d
s
i
m
u
l
at
i
o
n
s
an
d
y
i
el
d
-
g
a
p
an
al
y
s
i
s
Pat
r
i
ci
o
G
ra
s
s
i
n
i
a
,
L
e
n
n
y
G
.J. van
B
u
s
s
el
,
J
u
s
t
i
n
V
an
W
art
a,
J
o
o
s
t
W
o
l
f,L
i
ev
en
C
l
ae
s
s
e
n
s
,
d
,
H
ai
s
h
u
n
Y
an
g
a,
H
en
d
r
i
k
Bo
o
g
aar
d
e,
H
u
g
o
d
e G
ro
o
t
e,Mar
t
i
n
K
.
v
a
n
It
t
ers
u
m
b
,
K
en
n
e
t
h
G
.
Cas
s
m
an
/
2
0
1
5
Pres
e
n
t
ed
a
c
as
e
s
t
u
d
y
(N
eb
ras
k
a
-
U
SA
an
d
a
t
a
n
a
t
i
o
n
a
l
s
c
al
e
f
o
r
A
rg
en
t
i
n
a
a
n
d
K
en
y
a)
o
n
t
h
e
a
p
p
l
i
cat
i
o
n
o
f
an
ex
p
l
i
c
i
t
ra
t
i
o
n
a
l
e
d
e
s
i
g
n
ap
p
ro
ach
i
n
i
d
e
n
t
i
fy
i
n
g
t
h
e
d
at
a
s
o
u
rc
es
w
h
i
c
h
s
i
m
u
l
at
e
s
Cr
o
p
(
m
ai
ze)
y
i
el
d
an
d
a
l
s
o
h
el
p
s
i
n
q
u
a
n
t
i
fy
i
n
g
t
h
e m
ai
ze y
i
el
d
g
ap
s
.
Su
g
g
e
s
t
ed
t
h
e
ro
b
u
s
t
g
u
i
d
e
l
i
n
e
s
fo
r
an
a
l
y
zi
n
g
t
h
e
cr
o
p
y
i
el
d
g
ap
s
,
access
i
n
g
t
h
e
cl
i
m
at
e
an
d
l
a
n
d
u
s
e
c
h
a
n
g
e
s
a
t
g
l
o
b
a
l
l
e
v
e
l
t
o
a
d
d
r
es
s
t
h
e
i
s
s
u
e
s
of cr
o
p
y
i
el
d
.
[1
3
]
9.
Pred
i
c
t
i
o
n
o
f cr
o
p
y
i
el
d
u
s
i
n
g
b
i
g
d
at
a
W
u
Fan, Che
n
C
h
o
n
g
,
G
u
o
X
i
a
o
l
i
n
g
,
Y
u
H
u
a
W
an
g
J
u
y
u
n
/
2
0
1
5
D
ev
el
o
p
e
d
a
n
o
v
el
m
o
d
el
i
.e
N
eare
s
t
n
e
i
g
h
b
o
r
s
m
o
d
e
l
i
n
g
t
o
c
al
c
u
l
at
e
an
d
pre
d
i
ct
t
h
e y
i
el
d
o
f cr
o
p
d
e
p
e
n
d
s
o
n
t
h
e av
a
i
l
ab
l
e B
i
g
d
a
t
a s
et
s
.
[7
]
10.
T
h
e u
s
e o
f sa
t
e
l
l
i
t
e d
at
a f
o
r
cro
p
y
i
el
d
g
a
p
an
al
y
s
i
s
D
av
i
d
B.
L
o
b
el
l
/
2
0
1
3
D
i
s
c
u
s
s
e
d
t
h
e
u
s
e
o
f
rem
o
t
e
s
e
n
s
i
n
g
t
ech
n
o
l
o
g
y
t
o
i
d
e
n
t
i
fy
an
d
m
eas
u
re
t
h
e
ca
u
s
e
s
o
f
y
i
el
d
g
a
p
s
a
n
d
t
h
e
as
s
es
s
t
h
e
i
m
p
act
o
n
t
h
e
o
v
era
l
l
cr
o
p
y
i
el
d
.
Rep
o
rt
ed
v
ery
s
i
m
p
l
e
m
et
h
o
d
o
l
o
g
i
e
s
t
o
m
eas
u
re
t
h
e
y
i
el
d
d
i
ffere
n
ce
w
i
t
h
res
p
e
ct
t
o
s
ea
s
o
n
,
en
v
i
ro
n
m
en
t
an
d
t
h
e l
an
d
u
s
e.
[1
4
]
11.
Y
i
el
d
g
ap
a
n
a
l
y
s
i
s
w
i
t
h
l
o
c
al
t
o
g
l
o
b
a
l
rel
ev
a
n
ce
-
A
rev
i
ew
Mart
i
n
K
.
v
an
I
t
t
ers
u
m
a,
K
en
n
et
h
G
.
Cas
s
m
an
b
,
Pat
r
i
ci
o
G
ra
s
s
i
n
i
b
,
J
o
o
s
t
W
o
l
fa,
Pa
b
l
o
T
i
t
t
o
n
e
l
l
,
Z
v
i
H
o
c
h
m
an
d
D
i
s
c
u
s
s
e
d
ab
o
u
t
t
h
e
v
ari
o
u
s
m
et
h
o
d
s
u
s
e
d
o
n
q
u
an
t
i
fy
i
n
g
t
h
e
y
i
el
d
g
a
p
s
a
t
l
o
ca
l
-
to
-
g
l
o
b
a
l
rat
i
o
.
Rep
o
rt
ed
few
s
t
a
n
d
ard
o
p
era
t
i
o
n
m
et
h
o
d
s
,
em
p
l
o
y
ed
i
n
q
u
a
n
t
i
fy
t
h
e
cro
p
y
i
el
d
p
o
t
en
t
i
al
o
n
t
h
e
d
at
a
co
l
l
e
ct
ed
fro
m
t
h
e
farm
ers
o
f
w
es
t
e
rn
K
en
y
a,
N
eb
ras
k
a (U
S
A
) and
V
i
ct
o
r
i
a (A
u
s
t
ral
i
a).
St
u
d
y
reco
m
m
en
d
e
d
fo
r
t
h
e
u
s
e
o
f
acc
u
rat
e
a
n
d
c
u
rre
n
t
y
i
el
d
d
a
t
a,
w
i
t
h
ca
l
i
b
ra
t
e
d
an
d
v
al
i
d
a
t
ed
cro
p
m
o
d
el
s
an
d
u
p
s
c
al
i
n
g
m
et
h
o
d
s
i
n
t
h
e
p
re
d
i
ct
i
o
n
o
f cr
o
p
y
i
el
d
.
[1
5
]
3.
CONCL
US
I
O
N
As
a res
ult of
pe
netrati
on of technolo
gy into agric
ultur
e
fiel
d,
the
re is a m
a
rg
i
nal i
m
pr
ove
m
ent in th
e
pro
du
ct
ivit
y.
The
i
nnovat
io
ns
ha
ve
le
d
to
ne
w
c
on
ce
pts
li
ke
di
gital
agr
ic
ultur
e,
sm
art
fa
rm
ing
,
preci
sio
n
agr
ic
ultur
e
et
c
.
In
t
he
li
te
ratur
e
,
it
has
be
e
n
obse
rv
e
d
t
hat
analy
sis
ha
s
been
done
on
a
gr
ic
ultur
e
so
il
s,
hidden
patte
r
ns
disco
ver
y
us
ing
data
set
relat
ed
to
cl
i
m
at
i
c
conditi
on
s
a
nd
c
rop
yi
el
ds
data.
Th
e
act
ivit
ie
s
of
agr
ic
ultur
e
fiel
d
are
num
ero
us
li
ke
weat
her
forecast
in
g,
s
oi
l
qu
al
it
y
assessm
ent,
seeds
s
el
ect
ion
,
cr
op
yi
el
d
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Agricultu
re
Data A
na
ly
ti
cs in
Crop Yie
ld
Esti
ma
ti
on:
A Cri
ti
cal Revi
ew
(
B
M Sa
ga
r
)
1093
pr
e
dicti
on
et
c
.
In
t
his
s
urvey,
the
s
pecific
act
ivit
y,
crop
yi
el
d
pr
e
dicti
on
ha
s
bee
n
s
ur
veyed
a
nd
the
m
ajo
r
tren
ds
hav
e
b
e
en
ide
ntifie
d.
The
rice
crop
yi
el
d
pr
edict
ion
has
bee
n
done
in
the
s
ta
te
of
Ma
harashtra
us
i
ng
da
ta
m
ining
te
chn
iq
ues
in
on
e
of
the
wor
ks
[
8].
The
a
na
ly
sis
has
been
done
us
in
g
m
achine
le
arn
i
ng
fr
am
ewo
r
k
WE
KA.
In
th
e
w
ork
ca
rr
ie
d
out
in
[
9]
,
var
io
us
al
gorithm
s
app
li
ed
in
the
assess
m
ent
crop
yi
el
d
an
d
m
echan
ism
for
knowle
dge
dis
cov
e
ry
has
bee
n
disc
us
se
d.
T
he
ch
al
le
ng
es
and
oppo
rtu
niti
es
in
the
fiel
d
of
Bi
g
Data
a
na
ly
ti
c
s
in
ag
ricult
ure
has
been
disc
usse
d
in
[
6]
with
a
case
st
ud
y
of
Nethe
rlan
ds.
Fu
zzy
l
og
ic
de
sign
s
ha
ve
be
en
use
d
in
opti
m
iz
ing
t
he
c
rop
yi
el
ds
and
the
sam
e
has
bee
n
e
xp
la
ined
i
n
the
res
earch
w
ork
i
n
[5
]
.
A
cas
e
stu
dy
of
Nebras
ka
-
U
SA
a
nd
at
a
na
ti
on
al
scal
e
f
or
A
rg
e
ntina
a
nd
Ke
nya
has
bee
n
done
a
nd
pr
ese
nted
in
[
14
]
.
The
rem
ote
sensing
te
ch
nolo
gy
f
or
i
den
ti
fi
cat
ion
a
nd
m
easur
em
ent
of
the
ca
us
es
of
yi
el
d
gap
s
an
d
their
i
m
pact o
n final
cr
op yi
el
d
is p
resen
te
d
in
[1
5]
.
It
can
be
co
nclud
e
d
that
the
r
esearch
in
the
fiel
d
of
ag
ricul
ture
with
re
fere
nce
to
us
in
g
I
T
trend
s
li
ke
data
analy
ti
cs
is
in
it
s
infancy
.
As
the
foo
d
is
the
basic
need
of
hum
ans,
t
he
re
qu
i
r
e
m
ent
of
getti
ng
t
he
m
axi
m
u
m
yi
eld
s
us
in
g
opti
m
al
reso
urce
will
bec
om
e
t
he
necessit
y
in
near
f
uture
as
a
r
esult
of
grow
i
ng
popula
ti
on
.
T
he
s
urvey
ou
t
com
es
ind
ic
at
e
the
need
f
or
im
pr
ov
e
d
te
chn
i
qu
e
s
in
c
rop
yi
el
d
a
na
ly
ti
cs.
Ther
e
ex
ist
s a
l
ot of
researc
h s
cop
e
in
t
his res
earch
area
.
REFERE
NCE
S
[1]
Dhiv
y
a
B
H,
Ma
njul
a
R
,
Siv
a
Bh
ara
th
i
S,
Madhu
m
at
hi
R.
A Surve
y
on
Crop
Yield Predi
c
ti
on
b
ase
d
on
Agri
cul
tur
a
l
Data
,
Int
ernati
o
nal
Journal
of
In
novat
i
ve
Re
sear
ch
in
Sc
ie
nc
e, Enginee
ring
and
Technol
ogy
.
201
7;
6(3)
.
[2]
Jharna
Majumda
r,
Sneha
Nar
ase
e
y
app
a,
Shilpa
A
nkal
ak
i.
Ana
l
y
s
i
s
of
agr
ic
u
lt
ure
dat
a
using
d
atam
ini
ng
t
ec
hniqu
es:
appl
i
ca
t
ion
of
bi
g
data.
Journal
o
f
B
ig
data
.
2017
.
[3]
Majumdar
J,
A
nkal
ak
i
S.
Com
parison
of
cl
ust
ering
algorit
hms
using
quali
ty
metric
s
wit
h
inv
ariant
fe
ature
s
ex
trac
te
d
from p
lant
l
eaves.
In
te
r
nat
ion
al
Conf
erence
on
Com
putationa
l
Sci
ence
a
nd
Engi
n
ee
ring
.
2016.
[4]
D
Ramesh,
B
Vi
shnu
Vardha
n.
Data
Mining
Techni
ques
and
Applicati
ons
to
Agric
ult
ur
al
Yi
el
d
D
at
a
.
Int
ernati
ona
l
Journal
of
Ad
va
nce
d
Re
search
i
n
Computer
and
Comm
unic
ati
on
Engi
ne
ering.
20
13
;
2(9).
[5]
Sw
aru
pa
Rani
.
The
Im
pac
t
of
Data
Anal
y
t
ic
s
i
n
Crop
Mana
ge
m
ent
base
d
on
W
ea
the
r
Condi
tions
.
Inte
rna
ti
on
al
Journal
of Engin
ee
ring
T
ec
hnolo
g
y
Scie
n
ce a
nd
Re
sea
rch
.
2017
; 4(5):299
-
308.
[6]
F
K
Van
Eve
rt,
S
Fountas,
D
Ja
kovet
i
c,
V
Crnoj
evi
c
,
I
Tra
vlos
,
C
Kem
pena
ar.
B
ig
Data
for
wee
d
cont
rol
and
cro
p
prote
c
ti
on.
John
W
il
e
y
&
Sons
Ltd on
beh
al
f
of E
urope
an
W
e
ed
R
ese
arc
h
Soci
ety
,
2017:
218
–
233.
[7]
W
u
Fan,
Chen
Chong,
Guo
Xiaol
ing
,
Yu
Hua.
Predi
ct
ion
of
cro
p
y
i
el
d
u
sing
Big
Data
.
8th
Inte
rna
ti
on
al
S
y
m
posium
on
Com
puta
ti
onal I
nte
lligen
ce a
nd
Design.
2015
.
[8]
Daksha
y
in
i
Pat
il,
M
.
S,
Shirdho
nkar
.
Ri
ce
Crop
Yiel
d
Predi
ctio
n
using
Data
M
ini
ng
T
ec
hniqu
e
s:
An
Overvi
ew
.
Inte
rna
ti
ona
l
Jou
rna
l
of
Advanc
e
d
Resea
r
ch in
C
om
pute
r
Scie
n
ce a
nd
Softwa
re E
ngine
er
ing, 2017
;
7(5):427
-
431
.
[9]
Dhiv
y
a
B
H,
Ma
njul
a
R
,
Siv
a
Bh
ara
th
i
S,
Madhu
m
at
hi
R.
A Surve
y
on
Crop
Yield Predi
c
ti
on
b
ase
d
on
Agri
cul
tur
a
l
Data
,
Int
ern
ation
al
Journa
l
of
Inn
ovat
iv
e
Rese
arch
in
Sc
ie
n
ce
,
En
gine
er
ing
and
T
e
chnol
og
y
.
2017;
6(3):4177
-
4182.
[10]
Yogesh
Gandge
,
Sandh
y
a.
A
Stud
y
on
Vari
ous
Data
Mining
T
e
chni
ques
for
Cr
op
Yiel
d
Pred
ict
ion,
In
te
rna
ti
on
a
l
Confer
ence
on
El
e
ct
ri
ca
l
,
Elec
tr
onic
s,
Com
m
unic
ation,
Com
puter
and
Optimiza
tion
Te
chni
qu
es,
IEE
E
,
2017;420
-
423
[11]
R.
Suja
tha,
P.Isa
kki
Devi
.
A Stud
y
on
Crop
Yiel
d
Forec
asti
ng
Us
ing
Cla
ss
ifica
t
ion Te
chn
ique
s
,
IEEE,
2016
.
[12]
V.
Sellam
and
E
.
Poovam
m
al
.
Pr
edi
c
ti
on
of
Crop
Yiel
d
using
R
e
gre
ss
ion
Anal
y
si
s,
India
n
Journal
of
Sci
ence
and
Te
chno
log
y
,
201
6;
9(38).
[13]
Patri
ci
o
Grass
in
ia
,
L
enn
y
G.J.
v
an
Buss
el
,
Jus
ti
n
Van
W
art
a,
Joos
t
W
olf,
Li
ev
en
Cla
essens,
d,
Haishun
Yanga
,
Hendri
k
Booga
a
rde
,
Hugo
de
Gr
oote
,
Marti
n
K.
van
It
te
rsum
b,
Kenne
th
G
.
Cas
sm
an.
How
good
is
good
enoug
h
?
Data
r
equi
r
ements
for
r
el
i
abl
e
c
r
op
y
ield
sim
ula
t
i
ons a
nd
y
i
el
d
-
g
a
p
anal
y
s
is
.
Fi
el
d
Crops Re
sea
rch
.
2015;
49
–
63
.
[14]
David
B.
Lob
el
l
,
The use
of
sat
ellit
e
da
ta for crop
y
ield
g
ap
an
aly
s
is,
Fie
ld
Crops R
ese
arc
h
-
143
,
20
13;
56
–
64.
[15]
Marti
n
K.
van
I
t
te
rsum
a,
Kenne
t
h
G.
Cassm
anb,
Patri
ci
o
Gr
assini
b
,
Joos
t
W
olfa
,
Pablo
Ti
t
tone
l
l,
Zvi
Hochm
and.
Yiel
d
g
ap ana
l
y
s
is wi
th loc
a
l to
g
loba
l
releva
n
ce
-
A re
vie
w
.
Fie
ld Crops
Resea
rch
–
143,
2013;
4
–
17
Evaluation Warning : The document was created with Spire.PDF for Python.