Philadelphia University

Department of Communication & Electronics

Engineering

Probability & Random Variables

Chapter Five

Instructor
Ibrahim N. Abu-Isbeih

Email: [email protected]
Website: www.abusbeih.com/ecourse
 Ch.5: Operations on Multiple
Random Variables
1. Expected Value of a Function of Random Variables
2. Joint Moments
3. Joint Characteristic Functions

Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 2

1: Expectation
• The expected value of a function of random variables X and Y
is given by
 
g  E[ g ( X , Y )]    g ( x, y ) f X ,Y ( x, y )dxdy
 

• For N random variables X1 , X2 , ….. XN

g  E[ g ( X 1 ,  , X N )]
 
    g ( x1 ,  , xN ) f X1 ,, X N ( x1 ,  , xN )dx1    dx N
 

Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 3

2: Joint Moments
Joint Moments about the Origin:
• The joint moments of the random variables X and Y about the
origin are defined by
 
mnk  E[ X Y ]   n k
 x n y k f X ,Y ( x, y )dxdy
 

The sum n+k is called the order of the moments.

Clearly:
mn 0  E[ X n ] are the moments of X
m0 k  E[Y k ] are the moments of Y

Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 4

 The first order joint moments:
m10  E[ X ]  X the mean value of X
m01  E[Y ]  Y t he mean value of Y
The second order joint moments:

m20  E[ X 2 ] the mean square value of X

m02  E[Y 2 ] the mean square value of Y
m11  E[ XY ] the correlatio n of X and Y

Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 5

 The correlation of X and Y is important to later work:
 
RXY  m11  E[ XY ]    xy f X ,Y ( x, y )dxdy
 

• If RXY  E[ X ]E[Y ]
then X and Y are said to be uncorrelated.

• Statistical independence of X and Y is sufficient to guarantee

they are uncorrelated but the converse is not necessarily true
in general.

• If R XY  0
then X and Y are called orthogonal.

Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 6

Example: Let X be a random variable that has a mean value E[X]=3
and variance  X2  2 . Another random variable is defined by
Y  6 X  22
Find the mean value of Y, the variance of Y and the correlation of
X and Y.
Solution: 2 2 2
 X  E[ X ]  X
2
 E[ X ]   X  X  2  32  11
2 2

E[Y ]  E[6 X  22]  6 E[ X ]  22  4

E[Y 2 ]  E[( 6 X  22) 2 ]  E[36 X 2  264 X  484]
 36 E[ X 2 ]  264 E[ X ]  484  88
2
  E[Y ]  Y  72
2
Y
2

RXY  E[ XY ]  E[6 X 2  22 X ]  6 E[ X 2 ]  22 E[ X ]
 6(11)  22(3)  0
Since RXY  0, X and Y are orthogonal
RXY  E[ X ]E[Y ], X and Y are not uncorrelat ed
Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 7
 Joint Central Moments:
• The joint central moments of the random variables X and Y
are defined by
 nk  E[( X  X ) n (Y  Y ) k ]
 
  ( x  X ) n ( y  Y ) k f X ,Y ( x, y )dxdy
 

The sum n+k is called the order of the moments.

The first order central moments:

10  E[ X  X ]  0
01  E[Y  Y ]  0

Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 8

 • The second order central moments:
 20  E[( X  X ) 2 ]   X2
02  E[(Y  Y ) 2 ]   Y2
11  E[( X  X )(Y  Y )]  C XY
• The second order moment μ11 is called the covariance of X
and Y and is given by
C XY  11  E[( X  X )(Y  Y )] 
11 C
 XY
 20 02  X  Y
 RXY  E[ X ]E[Y ]
is called the correlatio n coefficien t
• If X and Y are either independent or uncorrelated then
CXY = 0
• If X and Y are orthogonal then
C XY   E[ X ]E[Y ]
Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 9
Example:
From the previous example
X 3  X2  2
Y  6 X  22
Y 4 R XY  0
Find the covariance of X and Y.
Solution:

C XY  RXY  X Y  12

Note that
C XY   X Y , because X and Y are orthogonal

Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 10

3: Joint Characteristic Functions
• The joint characteristic function of two random variables X
and Y is defined by
 X ,Y (1 , 2 )  E[e j1 X  j2Y ]
 
  e j1x  j2 y f X ,Y ( x, y )dxdy
 

The joint moments can be found as follows

 n  k  X ,Y (1 , 2 )
mnk  ( j ) n  k
1n 2k
1 0,2 0

 X (1 )   X ,Y (1 ,0) and Y (2 )   X ,Y (0, 2 )

are the marginal characteri stic functions
Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 11
 Example:
Two random variables X and Y have the joint characteristic function
212 822
 X ,Y (1 , 2 )  e

Find X , Y , RXY and C XY

Solution:

1 X ,Y (1 , 2 )  212 822

X  m10  ( j ) 1 0
  j (41 )e 0
1 
1 0
2
1 0,2 0
1 0,2 0

1 X ,Y (1 , 2 )  212 822

Y  m01  ( j ) 0 1
  j (162 )e 0
1 
0 1
2
1 0,2 0
1 0,2 0

 2
 X ,Y (1 , 2 ) 2 2
RXY  m11  ( j )11   (41 )( 162 )e 21 82 0
11 12 1 0,2 0
1 0,2 0

C XY  RXY  X Y  0  X and Y are uncorrelat ed

Instructor: Ibrahim Abu-Isbeih Probability & Random Variables - CH.5 12