Professional Documents
Culture Documents
,
( )
2 2
( )
2
.
y
f b y u u =
Define,
( ) ( ) ( )
( ) ( ) ( ) ( ) ( ) ( ) ( ) ( )
, 1 2 1 1 2 2
1 1 2 2
, , , , , ,
1
X Y
P x y P x P y
g x E g X g y E g Y
u u o u u
o
=
(
+
, (2.1)
where, ( )
1
g x and
2
( ) g y are bounded function on
2
9 .
note that We ( )
1 2
, ,
,
,
X Y
P x y,u u o
is a proper bivariate
distribution for a suitable choice of o e9. Based on the
distribution (2.1), in the following ntroduce three
types of BZIPSD.
Type-I BZIPSD
we i
: When there is an inflation only at
( x ) ( ) , 0, 0 y
' '
= , we define the BZIPSD as
( )
( ) ( ) ( ) ( )
( ) ( ) ( )
, 1 2
1 2
, 1 2
1 0, 0, , , , , 0, 0 , 0 1
, , , , ,
, , , , , , , 0, 0
X Y
X Y
P x y
P x y
P x y x y
u u o
u u o
u u o
' '
+ = < <
=
' '
(2.2)
Type-II BZIPSD: When there is inflation at X component only, we define the BZIPSD as
(2.3)
Type-III BZIPSD: When there is inflation at component only, we define the BZIPSD as
(2.4)
In the present discussion we focus only on Type-IBZIPSD,
re
Moment Generating Function
of (X,Y) is
( )
( ) ( ) ( ) ( )
( ) ( ) ( )
, 1 2
1 2
, 1 2
1 0, , , , , , 0, 0 , 0 1
, , , , ,
, , , , , , , 0, 0
X Y
X Y
P y x y
P x y
P x y x y
u u o
u u o
u u o
' '
+ = < <
=
' '
Y -
( )
( ) ( ) ( ) ( )
( )
, 1 2
1 2
, 1 2
1 , 0, , , , , 0, 0 , 0,1, 2,
, , , , ,
, , , , , , 0,1, 2, , 1, 2, 3,
0 1
X Y
X Y
P x x y x x
P x y
P x y x y
u u o
u u o
u u o
' '
+ = =
=
= =
< <
sults on the remaining two can be obtained analogously. The moment generating function
( ) ( )
( )
( ) ( ) ( ) ( ) ( )
{ }
( ) ( ) ( )
{ } { } { }
1 2
1 2
, 1 2
, 1 2
1 1 2 2 1 1 1 1 2 2 2 2
,
,
1
t X t Y
X Y
X Y
t X t Y
M t t E e
M t t
M t M t E e g X M t E g X E e g Y M t E g Y o
+
=
( ( = + + ( (
(2.5)
Therefore, we have
( ) M t M ( )
( ) { }
( ) ( )
( ) { }
1 , 1
1 1
2 , 2
2 2
, 0
1
0,
1
X X Y
Y X Y
t
M t
M t M t
M t
= +
=
= +
=
(2.6)
where ( )
1 X
M t
( )
j
f u ' and (
j
f ) u ''
( )
j
f u
u
c
c
denote
and ( )
2 Y
M t
m vari
are the moment generat-
ing fun f rando ables ctions o X and Y of zero-
inflated power series distribution nd a ( )
1 1
M t and
( )
2 2
M t are the moment generating functions dom
s having power series distribution with parame-
ters
1
of ran
variable
u and
2
u
respectively.
Su ose pp and
( )
2
2
j
f u
u
c
c
ives us
respectively for This g
( ) (
( )
1, 2 j = .
)
( )
1 1 1
1
f
E X M
f
1
0
X
u u '
' = =
u
( ) ( )
( )
( )
2 2 2
2 2
0
Y
f
E Y M
f
u u
u
'
' = =
( )
( )
( )
( )
( )
2
1 1 1 1
1 1 1 1 1
1 1 1 1
( )
f
Var X f f
f f
u u tu
u u u
u u
| |
'
'' ' | = +
|
\ .
Copyright 2011 SciRes. AM
P. M. KRISHNA ET AL. 826
( )
( )
( )
( )
( )
2
2 2 2 2
2 2 2 2 2
2 2 2 2
( )
f
Var X f f
f f
u u tu
u u u
u u
| |
'
'' ' | = +
|
\ .
,
and the correlation coefficient is
( ) ( )
( ) ( ) ( )
( )
( ) ( ) ( )
( )
( ) ( )
( )
( )
( ) ( )
( )
( )
1 1 1 1 1 1 2 2 2 2 2 2
1 1 2 2
1 2
2 2
1 1 2 2
1 1 1 2 2 2
1 1 1 1 1 2 2 2 2 2
1 1 2 2
f e f e f f e f e f
e f e f
f f
f f
f f f f
f f
u u u u u u
u u
u u
o
u u
u u u u
u u u u u u
u u
| |
| ' || ' |
|
| |
| | |
\ .\ .
|
=
|
| | | |
' '
| ' '' '
'' | | + +
| |
|
\ . \ .
\ .
(2.7)
3. Estimation of the Parameters of BZIPSD
Let ,
( ) , 1, 2, 3,
i i
X Y i n = be a random sample ob-
served from BZIPSD ( )
1 2
, , , . u u o The likelihood
funct e is given by. ion for the observed random sampl
( )
( ) ( ) ( )
( ) ( )
1 2
1
, , , ; , ,
i
a
n
L x y u u o
= (3
1 2
1
1 2
1 0, 0, , ,
, , , ,
i
XY
i
a
XY i i
P
P x y
u u o
u u o
=
+
[
.1)
where if 0 and otherwise.
The co din unct ven by,
1
i
a =
rrespon
( ) ( ) , 0,
i i
x y =
g log likelihood f
0
i
a =
ion is gi
( )
( ) ( ) ( )
1 2
0 1 2
lo ,
log 1 0, 0, , ,
XY
n P
u u
u u o
( )
, 1 2
1 1
g , , ; , ,
log log , , , ,
n n
i i X Y i i
i i
L x y
a a P x y
o
u u o
= =
+ +
) = + (3.2
log
0
L c
=
c
,
1
log
0
L
u
c
=
c
,
2
log
0
L
u
c
=
c
and
log
0
L
o
c
=
c
give the following equations.
( ) ( )
) (
0 , 1 2
0, 0, , ,
(1 ) 0, 0, ,
X Y
a
n P
P
u u o
u u +
1
, 1 2
1
0
,
n
i
i
X Y
o
=
+ =
(3.3)
( )
( )
( ) ( )
( )
( )
( )
1
1
0 , 1 2
, 1 2
, 1 2
1
, 1 2
0, 0, , ,
1 0, 0, , ,
, , , ,
0
, , , ,
X Y
X Y
n
X Y i i
i
i
X Y i i
n P
P
P x y
a
P x y
u
u
u u o
u u o
u u o
u u o
=
+
|
+ | =
\ .
|
|
(3.4)
( )
( )
( )
( )
( )
( )
2
2
0 , 1 2
, 1 2
, 1 2
1
, 1 2
0, 0, , ,
1 (0, 0, , , )
, , , ,
0
, , , ,
X Y
X Y
n
X Y i i
i
i
X Y i i
n P
P
P x y
a
P x y
u
u
u u o
u u o
u u o
u u o
=
+
| |
+ | =
|
\ .
( )
( )
( )
( )
( )
( )
0 , 1 2
, 1 2
, 1 2
1
, 1 2
0, 0, , ,
1 (0, 0, , , )
, , , ,
0
, , , ,
X Y
X Y
n
X Y i i
i
i
X Y i i
n P
P
P x y
a
P x y
o
o
u u o
u u o
u u o
u u o
=
+
| |
+ |
|
\ .
=
(3.6)
where denote
( )
,
(.)
X Y
P
o
,
(.)
X Y
P
o
c
c
Solving Equations (3.3) to (3.6) simultaneously we get
maximum likelihood estimators of the desired four pa-
rameters. We note that all the four likelihood equations
are non-linear in nature and do not have clo
lution. Now, we discuss a particular case
namely BZIPD.
sed form so-
of BZIPSD
4. Bivariate Zero-Inflated Poisson
Distribution
Let us set ( ) ( )
1
! a x x
= , ( ) ( )
1
! b y y
= , ( )
1
x
g x e
= ,
( )
2
y
g y e
= , ( )
1
1 1
f e
u
u
= , ( )
2
2 2
f e
u
u
= in the model
(2.1). Then we get BZIPD with probability mass func-
tion.
( )
)
( ) ( )
(3.5)
( )
( )
( )( ( )
( )( ) ( )
( ) ( )
( )
1 2
1 1 2
,
,
, 0, 0
1 ,
, 0, 0
where 1 1
x y
c x y
P
1 2 1 2
1 1 1 1
c c
2
( )
c
! !
X x Y y
x y
e
e e e
x y
c e
u
u u
o
=
e e e
u u u u
o
+
e
u u
u
+
x y
=
' '
+
=
=
(4. 1
The moment generating function of (X,Y) is
(4.2)
+ +
' '
( ) ( )
( ) ( ( ) ( )) ( ) ( ( ) ( ) ( ))
1 2
1 1 2 2
0 0
1 1 2 2
1 , ,
1 .
t x t y
x y
e P x e P y
g x E g X g y E g Y
u u
o
= =
= +
+
Copyright 2011 SciRes. AM
P. M. KRISHNA ET AL. 827
It is clear from the expressions of moment generating
functions of X and that the marginal distributions
of X and Y are univariate zero-inflated power series dis-
tributions with param ters
Y
e ( )
1
,u and ( )
2
,u respec-
tively. Further we have
( ) ( )
1 2
and E X E Y u u = =
( ) ( ) ( )
( ) ( ) ( )
1 1
2 2
1 1
and 1 1
Var X
Var Y
u u
u u
= +
= +
The correlation coefficient is turns out to be
( )
( )( ) ( )
( )
1 2
2 c u u
o u u
1 2
2
1 2 1 2
1 1 1
c e
u u u u
=
+ + +
(4.3)
Remark 1: When there is no inflation , the cor-
+
( ) 1 =
relation coefficient
is given by
( )
1 2
2
1 2
c
c e
u u
o u u
+
= ,
which coincides with the correlation coefficient given by
Laxminarayan et al. [8]
Remark 2: If we choose (.) g to be an her suit-
able bounded function, we will have different form of
BZIPD. Some other possible functions can be
( ) ,
z
g z a = 0 1 a ; < < ( )
z
g z e
=
0
etc.
Remark 3: If o = , we get Bivariate Zero-Inflated
Poisson distribution based on two independent random
variables.
Estimation of the Parameters of BZIPD
Suppose
( ) , ; 1, 2, ,
i i
x y i n =
is a random sample ob-
served from BZIPD
( ) , ; 1, 2, , i
1 2
, , n u u o =
. Then the
likelihood function is given by
( )
( )( ) ( ) ( )
( )
( )( ( )
1 2 1 2
1 2
1 2
1 2
1
( )
1
1 2
, , , ; ,
(1 ) 1 1 1
1
! !
i
i
i i
i i
n a
c c
i
a
x y
x y c c
i i
L x y
e e e
e
e e e e
x y
u u u u
u u
u u
u u o
t o
u u
o
)
+
=
+
=
+ +
| |
+ |
|
\ .
[
(4.4)
where 1
i
a = if ( ) ( ) , 0,
i i
x y = y ot 0 and otherwise.
The corresponding log likelihood is given by,
0
i
a =
( ) ( )
( )
( )( ) ( ) { }
( )
( ) ( ) ( )
1 2
1 1 1
log log log
i i i i i
i i i
a x a y a u u
= = =
+ +
( ) ( )( ) ( )
1 2 1 2
1 2
1 2 0 1 2
1 1
1 1
log , , , ; , log 1 1 1 1 log()
! log ! log 1
i i
n n
c c
i i
i i
n n n n n
x y c c
i i i i
i i
L x y n e e a a
x a y a e e e e
u u u u
u u
u u o o u u
o
+
= =
= =
= + + + +
+ +
(4.5)
The mles of the parameters can be obtained
equations
e
by solving
and
log
0
L
o
c
=
c
in the followi
simultaneously. These equations are given
ng:
log
0
L c
=
c
,
1 2
log
0
L
u
c
=
c
log
0
L
u
c
=
c
,
( )
( )( ) ( ) { }
)( ))
( )
( )
( (
1 2 1 2
1 2
1 1 1 e
u u
o
+
1 2
0
1
1 1 1
0
1
n
c c
i
i
c c
a
e
e e
u u u u
u u
o
+
=
+ + 1 n e e
+ =
(4.6)
+ +
( ) ( ) ( )( ) ( ) { }
( )
( )( ) { }
( )
( )
( )( )
1 2 2 1 1 2
1 2 1 2
1 2
1 2
( )
0
1
0
1
1
i=
\
1 1 1 1
1 1 1 1
0
1
i
i i
c c
c c
n
y c c
i i n
i
i
x y c c
n e c e e e e
e e e
a x
e e e
n n c a
e e e e
u u u u
u u u u
u u
u u
o o
o
o
u
o
+
+
=
+
+ +
| |
|
+ + =
|
+
.
c c u u
(4.7)
( ) ( ) ) ( )( ( ) { }
( )
( )( ) { }
( )
( )
( ) ( )
1 2 1 2 1
1 1
c
e e
u
+
2
1 2 1 2
2 1
1 2
( )
0
1
0
1
2
1 1
1 1 1 1
0
1
i
i i
c c c
c c
n
x c c
i i
n
i
i
x y c c
i
n e c e e
e e e
a y
e e e
n n c a
e e e e
u u u u
u u u u
u u
u u
o o
o
o
u
o
+
+
=
=
+ +
| |
|
+ + =
|
+
\ .
u
(4.8)
Copyright 2011 SciRes. AM
P. M. KRISHNA ET AL. 828
( ) ( )( ) { }
( )
( )( ) { }
( )( )
( )( )
1 2 1 2
1 2 1 2
1 2
1 2
( )
0
1
1 1
1 1 1 1
0
1
i i
i i
c c
c c
x y c c
n
i
x y c c
i
n e e e
e e e
e e e e
a
e e e e
u u u u
u u u u
u u
u u
o
o
+
+
=
+ +
| |
|
+ =
|
+
\ .
of the marginal distribution of Y to ZIPD we get Chi
square statistic = 0.6065 and P value = 0.4360. The table
value of = 3.841. Therefore, ZIPD fits well for
X
and Y dat
Thus now we can test whet
BZIPD
(4.9)
From the above equations, it is clear that Equations
(4.6) to (4.9) are non-linear in nature. Solving these
equations is computationally cumbersome. Laxminara-
yan et al. [7], adopt method of moments for the model
without inflation parameter (i.e. ). In their model
they have used estimates based on Method of Moment
Estimators (MME), which coincide with Maximum
Likelihood Estimators (MLE) of the marginal distribu-
tions. This is not the case for the joint distribution. We
have to solve four equations simultaneously in order to
et the MLEs. In the following we obtain maximum like-
lihood estimators for the following example and test for
goodness of fit.
5. An Application
The data set in Table 1 reported by Arbous and Kerrich
[10],
1 =
g
represents accidents sustained by 122 railway men
in consecutive periods of 6 and 5 years.
X is the accident distribution of 122 railway men dur-
ing 1937-1942 and Y is the accident distribution of 122
railway men during 1943-1947.
By assuming marginal distributions of X is ZIPD
( )
1
, . u The MLEs of X data are 0.8938 = and
1
1.3221 u =
.
Using these mles we fit the data of the mar-
ginal distribution of X to ZIPD
,
we get Chi square statis-
tic = 0.74843 and P value = 0.3869. If we fit the data
Table 1. Bivariate accident distribution of 122 railway men
0 1 2 3 4 5 6 7 Total
during two periods.
Y
X
0 21 14 8 1 44
1 17 12 8 3 1 1 42
9
5
2 6 9 2 2 2 21
3 1 1 3 3 1
4 1 3 4
2 2
6
7
Total 46 39 21 11 4 1 122
2
(1,0.05)
_
a.
her data is coming from
( )
1 2
, , . u u
n (4.5) using
m likeliho
0
Maximizing the log likelihood in the
Equatio MATLAB R12 software we get
maximu od estimators of the parameters as
0.94 = ,
1
1.210 u = ,
2
1.20 u = , 1.220 o =
-Inflated
. With
Zero Poisson
Distribution to the above data. The expected frequencies
are as shown in the Table 2.
From the chi-square good
calculated , is less than the table value of
alue is 0.392369. Hence we
o-Inflated Poisson Distribu-
tion fits well for the data.
Remark 4: There can be m ny ways to define k-vari-
A k-variate Zero-Inflated Power Series Distribution
ca b e ed
these parameters we fit Bivariate
ness of fit, we observed that
2
4.102062 _ =
9.488. The P v
hat Bivariate Zer
2
(4,0.05)
_ =
conclude t
a
ate ZIPSD by extending the above defined BZIPSD. One
of the ways is given below.
n e d fin as
( )
( ) ( )
( )
1 , , f 0
,
, f
X
P x
P x u
u o
i
,
P , x i x 0
X
x u o + =
where
=
( )
1 2
, , ,
k
u u u u
'
=
,
( )
1 2
, ,
k
X X X X =
( ) o
( ) ( ) ( ( ) { }
, ,
, ,
i
X
X i i
P x
P x g E g X
u
u o
| |
=
|
\ .
[ [
)
i
1 1 i=
k k
i
i=
x
( ) ,
i
X i i
P x u
ries Distri uti
is probability ma fu cti of ow r Se-
o
In ted to parameters involved in
m an a p si arly.
modat
ons. Further work under consideration is
te
model fo
2 Total
ss n on P e
b n.
ference
odel c
rela
be
the
ted
this
ttem mil
In the present work we introduced a new bivariate
zero-inflated power series distribution. This distribution
can accom e number of zero-inflated bivariate dis-
crete distributi
sting of independence for BZIPSD. Application of the
proposed r some other distributions like Bivari-
ate Zero-Inflated Negative Binomial Distribution or
k-variate zero inflated Poisson distribution can also be
Table 2. Expected frequencies using BZIPD.
Y
X
0 1
0 5500 8.7747 41.5161 21.1914 11.
1 11.6746 15.1344 14.5680 41.3770
2 8.9933 14.7627 15.3421 39.0981
Total 41.8593 41.4471 38.6848 121.9912
Copyright 2011 SciRes. AM
P. M. KRISHNA ET AL. 829
[1] L. Chin-Shang, K. Kyungmoo, J. P. Peterson and P. A.
Brinkley, Multivariate Zero-Inflated Poisson
and Their Applications, Technometrics, Vol. 41, No. 1,
99, pp. 29-38. doi:10.2307/1270992
doi: 10.1081/STA-120026576
considered. These models are useful to model zero-in-
flated bivariate data.
[5] P. Holgate, Estimation for the Bivariate Poisson Distri-
307/1269547
6. References
Models
19
[2] S. R. Deshmukh and M. S. Kasture, Bivariat
tion with Truncated Poisson Marginal Distributions,
Communication in Statistics: Theory and Metord
31, No. 4, 2002, pp. 527-534.
doi:10.1081/STA-120003132
e Distribu-
s, Vol.
[3] P. L. Gupta and R. C. Tripathi, Inflated Modified Power
Series Distributions with Applications, Communication
in Statistics: Theory and Metords, Vol. 24, No. 9, 1995,
pp. 2355-2374. doi:10.1080/03610929508831621
[4] R. L. Gupta and R. C. Tripathi Score Test for Zero-
Inflated Generalized Poisson Regression Model, Com-
munication in Statistics: Theory and Metords, Vol. 33,
No. 1, 2004, pp. 47-64.
bution, Biometrika, Vol. 51, No. 1-2, 1964, pp. 241-245.
[6] D. Lambert, Zero-Inflated Poisson Regression, with an
Application to Defects in Manufacturing, Technometrics,
Vol. 34, No. 1, 1992, pp. 1-14. doi:10.2
[7] J. Lakshiminarayana, S. N. N. Pandit and K. Srinivasa
Rao, On a Bivariate Poisson distribution, Communica-
tion in Statistics: Theory and Metords, Vol. 28, No. 2,
1999, pp. 267-276.
[8] M. K. Patil and D. T. Shirke, Testing Parameter of the
Power Series Distribution of a Zero-Inflated Power Series
Model, Statistical Methodology, Vol. 4, No. 4, 2007, pp.
393-406. doi:10.1016/j.stamet.2006.12.001
[9] M. K. Patil and D. T. Shirke, Tests for Equality of Infla-
tion Parameters of Two Zero-Inflated Power Series Dis-
tributions, Communications in Statistics: Theory and
Methods, Vol. 40, No. 14, 2011, pp. 2539-2553.
doi:10.1080/03610926.2010.489172
[10] A. G. Arbous and J. E. Kerrich, Accident Statistics and
the Concept of Accident Proneness, Biometrics, Vol. 7,
No. 1951, pp. 340-342. doi:10.2307/3001656
Copyright 2011 SciRes. AM