QDA for IE 2104690

Joint Probability Distributions

Jointly Distributed Random Variables

Expected Values
Covariance
Correlation

1
QDA for IE 2104690

Joint Probability Mass

Function
Two R.V.s, X and Y are observed simultaneously from one
experiment.
Two discrete random variables. For X and Y being discrete
R.V.s, the joint probability mass function

Note that joint PMF satisfies the probability model.

2
QDA for IE 2104690

Example
pXY(x, y) = (x+y)/30; x = 0,1,2,3; y = 0,1,2
Find P(X 2, Y = 1), P(X > 2, Y > 1), P(X > Y)

3
QDA for IE 2104690

Marginal PMF
Defined as the PMF of X alone or Y alone.
If X and Y are discrete R.V.s with joint PMF p(x,y), the
marginal PMF is obtained by summing p(x,y) for all y (or x).

Marginal PMF can be used to determine E[X] and E[Y].

From the previous example,

4
QDA for IE 2104690

Example
An electric display sign has 3 light bulbs in the 1st row and 4
light bulbs in the 2nd row.
X = # of bulbs in the 1st row that will be burned out at a
specified time t.
Y = # of bulbs in the 2nd row that will be burned out at a
specified time t.
Find P(X=Y) and P(X 2, Y 2)
Y
Find the marginal PMFs.
0 1 2 3 4
0 0.08 0.07 0.06 0.01 0.01
X 1 0.06 0.10 0.12 0.05 0.02
2 0.05 0.06 0.09 0.04 0.03
5
3 0.02 0.03 0.03 0.03 0.04
QDA for IE 2104690

Joint Probability
Density Function
For two continuous R.Vs, we define a joint PDF fXY(x,y) for
these two variables is a function satisfying

Then for any two-dimension set A

6
QDA for IE 2104690

Joint Cumulative
Distribution Function
Defn. A joint cumulative distribution function or
joint cdf of continuous R.Vs X and Y is defined as

y (x,y)

FXY(x,y) is the probability that X and Y

lies in this region.
x

7
QDA for IE 2104690

Examples (1.1)
A bank operates both a drive-up and a walk-up window. On a
randomly selected day, let
X = Proportion of time that the drive-up facility is used
Y = Proportion of time that the walk-up facility is used
The set of all possible values for (X, Y) is
D = {(x, y): 0 x 1, 0 y 1}

What is P (neither facilities is busy more than 1/4 of time)?

(Ans: 0.0109)
(0 X 1/4 and 0 Y 1/4 )
8
QDA for IE 2104690

Examples (2)
A randomly selected point (X, Y) in a unit square has the
uniform joint pdf given by

3 5
Find the joint CDF and P(X+Y 1)
2 4
1) x 0 or y 0, FX,Y(x,y) = 0
2) 0 x 1 and 0 y 1 1
x+y = 1

9
QDA for IE 2104690

Marginal PDF
If X and Y are continuous R.Vs. with joint PDF f(x,y)

Ex: f(x,y) = (21/4)x2y; x2 y 1. Determine the marginal pdf

of X and Y.

10
QDA for IE 2104690

Conditional Probability
Distribution
Let X and Y be two continuous RVs with joint pdf, f(x,y).
The conditional probability distribution of X given Y is
defined as

f XY ( x, y )
=
f X / Y ( x / y) , fY ( y ) > 0
fY ( y )

11
QDA for IE 2104690

Example (1.2)
Referring to the situation of banks facility, where
X = proportion of time that dive-up facility is busy and
Y = proportion of time that walk-up window is busy.
What is the conditional pdf of Y given that X = 0.8?

What is the probability that the walk-up facility is busy at most

half the time given that X = 0.8? (Ans: 0.39)

12
QDA for IE 2104690

Example (3)
Give the joint density function

x(1 + 3 y 2 )
, 0 < x < 2, 0 < y < 1
f XY ( x, y ) = 4
0
, elsewhere

x
find fX(x), fY(y) and fX/Y(x/y) (Ans: f X (=
x) , 0 <, x < 2
2 2
and determine (Ans: 3/64) 1+ 3y
=
fY ( y ) , 0 < y <1
1 1 1 2
P < X < Y =
4 2 3
f X / Y ( x / y=
)
x
, 0< x<2
2
13
QDA for IE 2104690

Independence of R.Vs
Definition: X and Y are independent iff

When X and Y are discrete

When X and Y are continuous

14
QDA for IE 2104690

Example 4
In an electric display problem, determine whether
X and Y are independent.
Y
0 1 2 3 4
0 0.08 0.07 0.06 0.01 0.01
X 1 0.06 0.10 0.12 0.05 0.02
2 0.05 0.06 0.09 0.04 0.03
3 0.02 0.03 0.03 0.03 0.04

15
QDA for IE 2104690

Example 5
Two independent measurements X and Y of rainfall are made
during a given period at a certain location. The pdf g(r) of
each measurement is

Determine P(X + Y 1)

16
QDA for IE 2104690

Ex. 6: Time before Series System

Failure
Let X and Y be two independent exponential R.V. with
parameters 1, 2. Find P(X < Y).

y=x

Two components connected in series

C1 C2

17
QDA for IE 2104690

Expected Value
For W = g(X, Y), E[W] can be determined directly from
pXY(x,y).

(X,Y) are discrete

(X,Y) are continuous

18
QDA for IE 2104690

Example 7
An instructor has given a short quiz consisting of 2 parts.
For a randomly chosen student, let X = the number of
points earned on the first part and Y = the number of points
earned on the second part. Suppose that the joint pmf of X
and Y is given in the table
a. If the score is the total number of
y points earned on the two parts,
p(x,y) 0 5 10 15 what is the expected record score
0 .02 .06 .02 .10 E(X+Y)?
b. If the maximum of the two scores
x 5 .04 .15 .20 .10
is recorded, what is the expected
10 .01 .15 .14 .01
recorded score?
19
QDA for IE 2104690

Example 8.1
A 1-lb mixed nut containing almonds, cashews, and peanuts.
The joint pdf of the amount X of almonds and amount Y of
cashews in a 1-lb can of nuts was

If 1 lb of almonds costs the company $1, 1 lb of cashews costs

$1.5 and 1 lb of peanuts costs $0.5, the total cost of the
contents of a can is

Determine the expected total cost. (Ans: $1.1) 20

QDA for IE 2104690

Covariance
When two rv X and Y are not independent, it is
frequently of interest to assess how strongly they are
related to one another

21
QDA for IE 2104690

Covariance
Covariance measures the degree that X and Y vary together.
If X,Y is positive
X-X and Y-Y tends to be positive or negative together.
X > X implies Y > X , or X < X implies Y < X
If X,Y is negative
X-X and Y-Y tends to have different signs.
X > X implies Y < X , or X < X implies Y > X
If X,Y = 0, X and Y are said to be uncorrelated.
If X and Y are independent, E[XY] = E[X]E[Y]. Therefore, X,Y = 0.
But X,Y = 0 does not imply independence.
Two dependent R.V.s can also have X,Y = 0.
22
QDA for IE 2104690

Properties of Covariance
Cov (X, X) = Var(X)
Cov (X, Y) = Cov(Y, X)
Cov (cX, Y) = c Cov(X, Y)
Cov (X, Y + Z) = Cov(X, Y) + Cov(X, Z)
Var(X+Y) = Var(X) + Var(Y) + 2 Cov(X, Y)

If Xis are independent

Cov[Xi, Xj] = 0, i j.
So, Var( Xi) = i Var(Xi) 23
QDA for IE 2104690

Example 9.1
The joint pmfs for X = automobile policy
deductible amount and Y = homeowner policy
deductible amount y
p(x,y) 0 100 200
Find COV(X,Y) x 100 .2 .1 .2
250 .05 .15 .3

24
QDA for IE 2104690

Example 8.2
The joint and marginal pdfs of X = amount of almonds
and Y = amount of cashews were

Find COV(X,Y) (Ans. -2/75)

25
QDA for IE 2104690

Correlation Coefficient
Let x and y be standard deviations of X and Y .

is called correlation coefficient or normalized covariance.

Toss a biased coin n times and observe # heads (X) and # tails
(Y). X + Y = n and E[X] + E[Y] = n.

26
QDA for IE 2104690

Correlation Coefficient
If a and c are either both positive or both negative
Corr(aX + b, cY +d) = Corr(X,Y)
For any two rvs X and Y, -1 Corr(X,Y) 1
If X and Y are independent, then = 0 but = 0 does not
imply independence.
If = 0 , X and Y are said to be uncorrelated.
= 1 or -1 iff Y= aX + b for some numbers a and b with a 0
A value of near 1 does not necessarily imply that increasing
the value of X causes Y to increase. It implies only that large X
values are associated with large Y values

27
QDA for IE 2104690

Example 9.2
The joint pmfs for X = automobile policy
deductible amount and Y = homeowner policy
deductible amount
y
Find Corr (X,Y) p(x,y) 0 100 200
x 100 .2 .1 .2
250 .05 .15 .3

28