Professional Documents
Culture Documents
PROBLEMS
93
7.5
Problems
6.3.2 Let X be a random variable with range [1, 1] and density function fX . Find (X ) and 2 (X ) if fX (x) = 0 for |x| > 1, and for |x| 1 Since the interval [1, 1] is symmetric about x = 0, any even density (f (x) = f (x)) will have = 0. Problem (a) is one. 3 (a) fX (x) = (1 x2 ), 4 = 3 x2 x4 3 x (1 x2 ) dx = [ ] 4 2 4 1 4
1 1 1 1
= 0.
1 1
E (X 2 ) =
3 3 x3 x5 x2 (1 x2 ) dx = [ ] 4 4 3 5 1 1 2 = E (X 2 ) 2 = . 5
1 5
Since fX
2 =
1
x2
1
cos(x/2) dx 4
1
22 [x sin(x/2) 4
1 1
2 2x sin(x/2) dx] 1
1
2 = 1 + x cos(x/2)
2 cos(x/2) dx]
=1
8 2
1 1
2 = . 6
1 1
E (X 2 ) =
1
x2
x3 x4 x+1 dx = + 2 6 8
2 = . 6
94
x(x2 + 2x + 1) dx =
1 1
34 1 = . 83 2 3 16 2 = . 8 15 5
E (X 2 ) =
x2 (x2 + 2x + 1) dx =
1
2 1 3 = . 5 4 20 6.3.3 The lifetime in hours of the ACME light bulb is a random variable T with density function 2 = E (X 2 ) 2 = fT (t) = 2 tet , = .05.
What are the expected value and variance for the lifetime? After integration by parts,
=
0
tfT (t) dt =
0
t exp(t) dt = 2
0
2 2
exp(t) dt = 2/.
Similarly, E (T ) =
0 2
2 t3 exp(t) dt = 6/2 ,
so 2 = E (T 2 ) 2 = 2/2. With = .05 we get = 40, 2 = 800. 6.3.6 Let T be a random variable with density fT (t) and range [0, ]. Find (T ) and 2 (T ) if fT (t) = 0 for t < 0, and for t 0 (a) fT (t) = 3e3t , This is just an exponential density with = 3, so integration by parts gives mu = 1/ = 1/3, 2 = 1/ 9 .
7.5. PROBLEMS
95
9t 3t e , 2 I think the factor two is an error. This should be (b) fT (t) = (b) fT (t) = 9te3t , making it a case of problem 6.3.3 with = 3. Thus = 2/ 3 , 2 = 2/ 9 . 3 . (1 + t)4
0
=
0
3 t dt = (1 + t)4
2
1 1 dt = , 3 (1 + t) 2
0
E (T ) =
0
3 dt = t (1 + t)4
1 dt = 1, (1 + t)2
so = 1/ 2 , 2 = E (T 2 ) 2 = 3 / 4 . 6.3.7 Let X be a random variable with density fX (x). Show that the function (a) = E ([X a]2 ), takes its minimum value when a = (X ), and in that case (a) = 2 (X ). We have
(x a)2 fX (x) dx =
This is a quadratic polynomial in a, and has its minimum when (a) = 0. Writing this equation we nd
0 = 2
xfX (x) dx + 2a
fX (x) dx,
or (X ) =
xfX (x) dx = a
fX (x) dx = a.
96
Of course by the denition of the variance 2 (X ) = (X ) = E ([X a]2 ). 6.3.17 Let X and Y be random variables. The covariance Cov (X, Y ) is dened by cov (X, Y ) = E ([X (X )][Y (Y )]). (a) Show that cov (X, Y ) = E (XY ) E (X )E (Y ). By linearity of the expected value, cov (X, Y ) = E ([X (X )][Y (Y )]) = E (XY ) (Y )E (X ) (X )E (Y ) + (X )(Y ) = E (XY ) E (X )E (Y ). (b) Using (a), show that cov (X, Y ) 0 if X and Y are independent. If X and Y are independent then E (XY ) = E (X )E (Y ), so cov (X, Y ) = E (XY ) E (X )E (Y ) = 0. (c) Show that V (X + Y ) = V (X ) + V (Y ) + 2cov (X, Y ). Using the denition, V (X + Y ) = E ([X + Y (X ) (Y )]2 ) = E ([(X (X ) + (Y (Y ))]2 ) = E ([X (X )]2 ) + 2E ([X (X )][Y (Y )]) + E ([Y (Y )]2 ) = V (X ) + V (Y ) + 2cov (X, Y ). 6.3.18 Let X and Y be random variable with positive variance. The correlation of X and Y is dened as (X, Y ) = (a) Using 17(c), show that 0V( X Y + ) = 2(1 + (X, Y )). (X ) (Y ) cov (X, Y ) . V (X )V (Y )
7.5. PROBLEMS = E( so by 17 (c), V( X Y X Y X Y + )=V( )+V( ) + 2cov ( , ) (X ) (Y ) (X ) (Y ) (X ) (Y ) = V (Y ) XY (X ) (Y ) V (X ) + + 2 E ( ) 2 2 (X ) 2 (Y ) (X ) (Y ) (X ) (Y ) =2+2 (b) Now show that 0 V( X Y ) = 2(1 (X, Y )). (X ) (Y ) E (XY ) (X )(Y ) = 2(1 + (X, Y )). V (X )V (Y ) XY (X ) (Y ) ) (X ) (Y ) (X ) (Y )
97
Use the result of (a) with X and Z = Y , noting that cov (X, Y ) = cov (X, Y ). (c) Show that 1 (X, Y ) 1. From (a) and (b) we get 0 (1 + (X, Y ))(1 (X, Y )) = 1 (X, Y )2 , or (X, Y )2 1, which is equivalent to 1 (X, Y ) 1.