0% found this document useful (0 votes)
66 views6 pages

Exercises 5

The document discusses order statistics and provides formulas for: 1) The probability density function and distribution function of the nth order statistic Y(n) of a random sample of size n from a uniform distribution on [0,θ]. 2) The expected value, variance, and probability density function of the kth order statistic Y(k) of a random sample of size n from a uniform distribution on [0,θ]. 3) The covariance and variance of the difference between two order statistics Y(k) and Y(j) of a random sample of size n from a uniform distribution on [0,θ].

Uploaded by

dan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
66 views6 pages

Exercises 5

The document discusses order statistics and provides formulas for: 1) The probability density function and distribution function of the nth order statistic Y(n) of a random sample of size n from a uniform distribution on [0,θ]. 2) The expected value, variance, and probability density function of the kth order statistic Y(k) of a random sample of size n from a uniform distribution on [0,θ]. 3) The covariance and variance of the difference between two order statistics Y(k) and Y(j) of a random sample of size n from a uniform distribution on [0,θ].

Uploaded by

dan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

ORDER STATISTICS Exercises

Answers

[6.74]

The probability density function and distribution function of Y are

1
f (y) = 0≤y≤θ
θ
y
F (y) = 0≤y≤θ
θ

y n

a. The distribution function of Y(n) is Fn (y) = θ 0≤y≤θ

ny n−1
b. The probability density function of Y(n) is fn (y) = θn 0≤y≤θ

c.
θ
yny n−1
Z
E(Y(n) ) = dy
0 θn
n 
= θ
n+1
Z θ 2 n−1
2 y ny
E(Y(n) ) = dy
0 θn
n  2
= θ
n+2
2
V (Y(n) ) = E(Y(n) ) − E 2 (Y(n) )
n  2 nθ 2
= θ −
n+2 n+1
2

=
(n + 1)2 (n + 2)

1
ORDER STATISTICS Exercises

[6.76]

a. The density function of Y(k) is

n! y k−1 θ − y n−k 1 
fk (y) =
(k − 1)!(n − k)! θ θ θ
k−1 n−k
n! y (θ − y)
= 0≤y≤θ
(k − 1)!(n − k)! θn

b. Z θ
E(Y(k) ) = yfk (y) dy
0
θ
y k (θ − y)n−k
Z
n!
= dy
0 (k − 1)!(n − k)! θn
Z θ
n! y k y n−k
= 1− dy
0 (k − 1)!(n − k)! θ θ
Z 1
n! y
= θ z k (1 − z)n−k dz with z=
(k − 1)!(n − k)! o θ
n! 1

(k − 1)!(n − k)! B(k + 1, n − k + 1)
k
= θ
n+1

c. Z θ
2
E(Y(k) ) = y 2 fk (y) dy
0
θ
y k+1 (θ − y)n−k
Z
n!
= dy
0 (k − 1)!(n − k)! θn
Z θ
n!θ y k+1 y n−k
= 1− dy
0 (k − 1)!(n − k)! θ θ
Z 1
n!θ2 y
= z k+1 (1 − z)n−k dz with z=
(k − 1)!(n − k)! o θ
n! 1
= θ2
(k − 1)!(n − k)! B(k + 2, n − k + 1)
k(k + 1)
= θ2
(n + 1)(n + 2)

2
ORDER STATISTICS Exercises

The variance of Y(k) is

2
V (Y(k) ) = E(Y(k) ) − E 2 (Y(k) )
k(k + 1) k2 θ2
= θ2 −
(n + 1)(n + 2) (n + 1)2
(n − k + 1)kθ2
=
(n + 1)2 (n + 2)

d.
E(Y(k) − Y(k−1) ) = E(Y(k) ) − E(Y(k−1) )
kθ (k − 1)θ
= −
n+1 n+1
θ
=
n+1

which is constant for all k and so the order statistics are equally spaced.

3
ORDER STATISTICS Exercises

[6.77]

a.

The joint density function of Y(j) and Y(k) is

n! yj j−1 h yk yj ik−1−j h yk in−k 1 2


f (yj , yk ) = − 1−
(j − 1)!(k − 1 − j)!(n − k)! θ θ θ θ θ

for 0 ≤ yj ≤ yk ≤ θ.

b. The covariance of Y(j) and Y(k) is given by

Cov(Y(j) , Y(k) ) = E(Y(j) Y(k) ) − E(Y(j) )E(Y(k) )

and
n!
E(Y(j) Y(k) ) =
(j − 1)!(k − 1 − j)!(n − k)!
Z θ Z yk
yj j−1 h yk yj ik−1−j h yk in−k 1 2
× yj yk − 1− dyj dyk
0 0 θ θ θ θ θ
n!
=
(j − 1)!(k − 1 − j)!(n − k)!
Z θ Z yk
yj  j h yk yj ik−1−j yk h yk in−k
× − 1− dyj dyk
0 0 θ θ θ θ θ
n!
=
(j − 1)!(k − 1 − j)!(n − k)!
Z θ
yk in−k yk yj j h yk yj ik−1−j
Z
yk h
× 1− − dyj dyk
0 θ θ 0 θ θ θ
n!θ
=
(j − 1)!(k − 1 − j)!(n − k)!
Z θ
yk in−k yk  1 zyk j h yk zyk ik−1−j
Z
yk h
× 1− − dzdyk
0 θ θ θ 0 θ θ θ
n!θ
=
(j − 1)!(k − 1 − j)!(n − k)!
Z θ
yk in−k 1 j
Z
yk k+1 h
× 1− z (1 − z)k−1−j dzdyk
0 θ θ 0

4
ORDER STATISTICS Exercises

Evaluating these integrals using the beta function gives

(k + 1)jθ2
E(Y(j) Y(k) ) =
(n + 2)(n + 1)

and so
(k + 1)jθ2 jθ  kθ 
Cov(Y(j) , Y(k) ) = −
(n + 2)(n + 1) n+1 n+1
2
j(n − k + 1)θ
=
(n + 1)2 (n + 2)

c.

The variance of Y(k) − Y(j) is

V (Y(k) − Y(j) ) = V (Y(k) ) + V (Y(j) ) − 2Cov(Y(k) , Y(j) )


h k(n − k + 1) j(n − j + 1) 2j(n − k + 1) i 2
= + − θ
(n + 1)2 (n + 2) (n + 1)2 (n + 2) (n + 1)2 (n + 2)
(k − j)(n − k + j + 1) 2
= θ
(n + 1)2 (n + 2)

5
ORDER STATISTICS Exercises

[6.91] The joint distribution of W(j−1) and W(j) is


n! n−j 1  −(wj +wj−1 )/θ
f (wj , wj−1 ) = (1 − e−wj−1 /θ )j−2 e−wj /θ e
(j − 2)!(n − j)! θ2
for 0 ≤ wj−1 ≤ wj < ∞.

Consider the transformation


Tj−1 = Wj−1
Tj = Wj − Wj−1
but identify Tj−1 as Y1 and Tj as Y2 for notational simplicity so that
Y1 = Wj−1
Y2 = Wj − Wj−1
wj−1 = y1
w j = y1 + y2


0 1
|J| =
1 −1
=1
and the joint probability density function of Y1 and Y2 is
n! j−2 −(y1 +y2 )/θ n−j 1  −(2y1 +y2 )/θ
f (y1 , y2 ) = 1 − e−y1 /θ e e
(j − 2)!(n − j)! θ2
and integrating over y1 gives the marginal probability density function of Y2 as
n! 1  −y2 /θ −y2 (n−j)/θ
f (y2 ) = e e
(j − 2)!(n − j)! θ2
Z ∞
j−2 −y1 /θ n−j+2
× 1 − e−y1 /θ e dy1
0
1 n!
= 2 e−y2 (n−j+1)/θ
θ (j − 2)!(n − j)!
Z 1
−θ
× (1 − u)j−2 un−j+2 du letting u = e−y1 /θ
0 u
1 n! Γ(j − 1)Γ(n − j + 2)
= e−y2 (n−j+1)/θ
θ (j − 2)!(n − j)! Γ(n + 1)
n − j + 1 −y2 (n−j+1)/θ
= e 0 < y2
θ
θ
so Wj − Wj−1 is exponential with mean n−j+1 .

You might also like