AMA2104 Probability and Engineering Statistics

3 Joint Distribution

Dr Bob He

1
Joint Distributions, discrete case

Definition
The function f(x, y) is a joint probability mass function of the discrete
random variables X and Y if
1. J ( x, y) > 0 for a II ( x, y)
2.

X y

3. P(X = x, Y = y) = f(x, y), where P(A, B) means P(A and B).

For any regin n in the x-y plane,
P((X, Y) En) = L J(x, y).
(x,y)EO

The definition can be extended to three or more random variables, e.g.,

J(x1, X2, ... , X n ),
2
 Example
We toss a die twice and let X and Y denote the number obtained
respectively. Find
0 IE[max(X, Y)], Var(max(X, Y))
G IE[X + Y], Var(X + Y)
We use the table to
summarize the
probability of
max(X, Y) taking
different values.
'X:.Y 1 2 3 4 5 6
1 1 2 3 4 5 6
2 2 2 3 4 5 6
3 3 3 3 4 5 6
4 4 4 4 4 5 6
5 5 5 5 5 5 6
6 6 6 6 6 6 6

3
Joint Distributions, continuous case

Definition
The function f(x, y) is a joint probability density function of the
continuous random variables X and Y if
1. f ( x, y) > 0 for aII ( x, y).
2.

1_: 1_: J(x, y)dxdy = 1.

3.

P((X, Y) En) = l f(x, y)dxdy,

for any region n in the x-y plane.

The definition can be extended to three or more random variables, e.g.,
f(x1, X2, ... , X n )-
4
Example
Let X and Y be two continuous random variables having the joint probability
density function

f ( X' y) ={ ¾ ( 2x
0,
+ 3y), 0 � X � 1, 0 � y � 1,
otherwise.

Verify that f is a joint PDF. Find P((X, Y) EA), where

½,
A= { (x, y) : 0 < x < ¼ < y < ½ }.

5
Double Integrals

Assuming that the region in R2 is a rectangle which we will denote

as follows, R = [a, b] × [c, d]. We will first divide up a 6 x 6 b
into n subintervals and divide up c 6 y 6 d into m subintervals.
This will divide up R into a series of smaller rectangles and from
each of these we will choose a point (xi , yj ).

6
Over each
of these smaller rectangles we will construct
a box whose height is given by f (xi , yj ).
Each of the rectangles has a base area of ∆A
and a height of f (xi , yj ) so the volume of each
of these boxes is f (xi , yj )∆A. The volume
under the function is then approximately,
n X
X m
V ≈ f (xi , yj ) ∆A.
i=1 j=1

To get a better estimation of the volume we

will take n and m larger and larger and to get the exact volume we
will need to take the limit.
n X
X m
V = lim f (xi , yj ) ∆A
n,m→∞
i=1 j=1

7
Definition
We define the double integral of the function f (x, y ) over the
rectangle R to be the limit
ZZ ZZ n X
X m
f (x, y )dA = f (x, y )dxdy = lim f (xi , yj ) ∆A.
R R n,m→∞
i=1 j=1

Iterated Integrals

ZZ Z b Z d Z d Z b
f (x, y )dA = f (x, y )dydx = f (x, y )dxdy
R a c c a

8
Bounded Irregular Region

Theorem
Suppose that f (x, y ) is continuous on the region R.
1 If R is a vertically simple region a 6 x 6 b,
y1 (x) 6 y 6 y2 (x), then
!
ZZ Z b Z y2 (x)
f (x, y )dA = f (x, y )dy dx.
R a y1 (x)

2 If R is a horizontally simple region c 6 y 6 d,

x1 (y ) 6 x 6 x2 (y ), then
!
ZZ Z d Z x2 (y )
f (x, y )dA = f (x, y )dx dy .
R c x1 (y )

9
Finding The Limits of Integration

1 Sketch the region of integration and label the bounding

curves.
2 Imagine a vertical line L cutting through R in the direction of
increasing y . Mark the y -values where L enters and leaves.
These are the y -limits of integration and usually functions of
x.

3 Choose x-limits that include all the vertical lines through R.

10
Example
The joint density for the random variables (X , Y ), where X is the temperature
change and Y is the proportion of the spectrum that shifts for a certain atomic
particle, is
(
10xy 2 , for 0 < x < y < 1,
f (x, y ) =
0 elsewhere.

11
Marginal distributions

Definition
The marginal distributions of X alone and Y alone are:
• For the discrete case,

g(x) = L f(x, y), and h(y) = L f(x, y)

X

• For the continuous case,

g(x) = 1-: f(x, y)dy, and h(y) = 1-: f(x, y)dx.

12
Example
Suppose that X and Y are two discrete random variables and their joint
probability mass function f(x, y) is given by the table below.

X
f(x,y) Row Totals
0 1 2
0 0.13 0.05 0.05 0.23
y 1 0.23 0.21 0.07 0.51
2 0.15 0.03 0.08 0.26
Column Totals 0.51 0.29 0.20 1.00

f(x,y) --1�-2
-0
- Row Totals
0.05 0.05
y

f(x,y) h(y)
g(x)

13
Example
Let X and Y be two continuous random variables having the joint
probability density function

f(x,y) = { t,c2x + 3y), Q :::; X :::; 1, Q :::; y :::; 1,

otherwise.

Find the marginal density functions g(x) and h(y).

14
Definition
Let X and Y be two random variables, discrete or continuous. The
conditional distribution of the random variable Y given that X = x is

f(x,y)
f(yIx) = provided that g(x) > 0.
g(x) '

Definition
Let X1 , ... , Xn be n random variables, discrete or continuous, with joint
probability distribution f(x1, ... ,xn ) and marginal distributions fi(x 1 ),
... , fn (xn ) respectively. Then we say that the random variables
X1 , ... , Xn are (mutually) independent if

Facts
Let X and Y be two independent random variables. Then

IE[XY] = IE[X]IE[Y].
15
Example
Let X and Y be two continuous random variables having the joint probability
density function
(
2
5
(2x + 3y ), 0 ≤ x ≤ 1, 0 ≤ y ≤ 1,
f (x, y ) =
0, otherwise.

Find E[X ], E[Y ], and E[XY ].

16
17
Moment Generating Functions (MGF's)
Definition
The moment generating function of a random variable X is a function
Mx(s) defined by

Example
Suppose X ~ Poisson(.X), find the MGF of X.

18
Theorem (properties of MGF)
Let random variables X and Y be independent. Let Mx(s) and My(s)
be th eir MGF respectively. Assume that they are finite at an open
interval that covers 0.
1. Mi_m ) (0) = IE(X m ) for every positive integer m.
2. Mx+y(s) Mx(s)My(s)
3. If Mx(s) My(s), then X and Y have the same distribution.
4. Max+b(s) = e bsMx(as) for any constants a and b.

19
20
Example
~ ~
Suppose X1 Poisson(.\1), X2 Poisson(.\2), ... , and
~
Xm Poisson(.\m ) are independent. Let Y = X1 + · · · + Xm . Then
~
Y Poisson(.\1 +···+A m ).

21
Example
Suppose X ~ Poisson(-\). Find JE[X] and Var(X).

22
Example
If X ~ Exp(A), find Mx(s). Find lE[X].

23
Example
Suppose X ~ N(O, 1). Find Mx(s).

24
25