Chapter

4. Con-nuous Random Variables

4.1 Continuous random variables
For con-nuous random variables, we focus on the
probability density func2on (pdf) f(x).

Defini-on: X is said to be a con2nuous random variable

if there exists a nonnega-ve func-on f, defined for all
real numbers x∈(−∞, ∞), having the property that
for any set B of real numbers, the probability that X
belongs to B is equal to:

f is called the pdf of X.

1
Continuous random variables

Since X must take on some value (in this course, it is
usually assumed that X cannot be infinite), the pdf
must sa-sfy:

The converse is also true: a nonnega-ve func-on f is the
pdf of some con-nuous r.v. X if and only if

2
Continuous random variables

All probability statements about X can be answered by
using the pdf, for example:





Note that the last line above is the cdf.

3
Continuous random variables

Rela-onship between pdf and cdf: The rela-onship
between the pdf and the cdf is as follows




 Continuous random variables

Interpreta-on of the pdf: Note that

when is small and f (.) is con-nuous at x=a. So the
probability that X is contained in an interval of length
ε around the point a is approximately .

5
 Continuous random variables

Example: The life-me in hours of a certain kind of radio tube

is a random variable having a probability density func-on
given by

What is the probability that exactly 2 of 5 such tubes in a
radio set will have to be replaced within the first 150
hours of opera-on? We assume that the events
t that the ith such tube has to be
replaced within this -me period, are independent.

6
More examples regarding the cdf
Example: which of the following func-ons are a cdf?

 More examples regarding the cdf

Example: The cumula-ve distribu-on func-on for some

random variable X is




and P(X=2)=1/2. What are the values of a and b?

 More examples regarding the cdf

Example: A con-nuous random variable X has the

following cdf:



We also know that P{X>1/2}=3/4. Find the values of a, b,
and c.
 More examples regarding the cdf

Example: The cdf of a r.v. is

Compute .

 Expecta-on and variance

Expecta-on and Variance

Defini-on: Recall that for a discrete random variable X,
the expected value is .
For a con-nuous random variable X with pdf f(x),

The variance of a con-nuous random variable is

11
 Expecta-on and variance

Proposi-on: If X is a con-nuous random variable with

pdf f(x), then for any real-valued func-on g,



Proposi-on: If a and b are constants, then

12
 Expecta-on and variance

Example: Find E[X] and Var[X] when the density func-on

of X is

13
 4.2 Special distribu-ons: uniform

Uniform distribu-on
A random variable is said to be uniformly distributed
over the interval if its pdf is given by

14
Special distribu-ons: uniform
Note: This is a valid pdf because

The cdf of a uniform random variable is

15
 Special distribu-ons: uniform

Example: Show that the expecta-on and variance of a

random variable that is uniformly distributed on
are equal to

16
 Special distribu-ons: uniform

Example: Buses arrive at a specified stop at 15-minute

intervals star-ng at 7am. That is, they arrive at 7,
7:15, 7:30, 7:45, and so on. If a passenger arrives at
the stop at a -me which is uniformly distributed
between 7 and 7:30, find the probability that he/she
waits less than 5 minutes for a bus.

Special distribu-ons: normal
Normal Random Variables
Defini-on: We say that X is a normal/Gaussian random
variable with parameters if the pdf of X is
given by

18
 Special distribu-ons: normal

(1) It can be checked that

but this requires some work.

(2) The central limit theorem jus-fies the use of the normal
distribu-on in many applica-ons. Roughly, the central
limit theorem says that if a random variable is the sum of
a large number of independent random variables, it is
approximately normally distributed.

(3) Some-mes normal distribu-on is used to model

logarithmic return:

19
 Special distribu-ons: normal

Linear transforma-ons of normal random variables:

An important fact about normal random variables is that
if X is normally distributed with parameters ,
then Y=aX+b is normally distributed with parameters

20
 Special distribu-ons: normal

A consequence of the preceding result is that if X is

normally distributed with parameters , then
the r.v. is normally distributed, with
parameters 0 and 1. Such a random variable is said to
have standard normal distribu2on.

21
 Special distribu-ons: normal

Mean and variance of normal random variables:

We start by finding the mean and variance of the
standard normal random variable . We
have


Since , , we obtain



22
Special distribu-ons: normal
cdf of a standard normal distribu-on:
One property of is that
Proof: Since the distribu-on of standard Gaussian r.v. is
symmetric , by the defini-on of
cdf, ,

23
 Special distribu-ons: normal

Suppose X is a normal r.v. with parameters .

Since has standard normal distribu-on,


That is, the cdf of a normal r.v. with parameters
is

24
 Special distribu-ons: normal

Example: If X is a normal random variable with

parameters , find
P(2<X<5).

25
 Special distribu-ons: normal

It is useful (for the Sta-s-cs part later) to memorize that

about 68% of values drawn from a normal distribu-on
are within one standard devia-on away from the
mean; about 95% of the values are within two
standard devia-ons; and about 99.7% lie within 3
standard devia-ons, i.e.


 Special distribu-ons: normal

Example: The army is developing a new missile. By observing

points of impact, launchers can control the mean of its
impact distribu-on. If the standard devia-on of the
impact distribu-on is too large, though, the missile will be
ineffec-ve. Suppose the Pentagon requires that at least
95% of the missiles must fall within 1/8 mile of the target
when the missiles are aimed properly. Assume the impact
distribu-on is normal. What is the maximum allowable
standard devia-on?

28
 Special distribu-ons: normal

Example: An expert witness in a paternity suit tes-fies

that the length (in days) of pregnancy is
approximately normally distributed with mean 270
and variance 100. The defendant in the suit is able to
prove that he was out of the country during a period
that began 290 days before the birth of the child and
ended 240 days before the birth. If the defendant
was, in fact, the father of the child, what is the
probability that the mother would have been the very
long or very short pregnancy indicated by the
tes-mony?
29
 Special distribu-ons: exponen-al
Exponen-al Distribu-on
Defini-on: A random variable with density func-on

is called an exponen-al random variable with parameter .
The cdf is

Remarks: an exponen-al random variable describes the -me
un-l a specific event occurs if the events occur according
to a Poisson process with rate : e.g., the amount of -me
un-l a telephone call you receive turns out to be a wrong
number.
30
 Special distribu-ons: exponen-al

Example: Suppose that the length of a phone call in

minutes is an exponen-al random variable with
parameter . If someone arrives immediately
ahead of you at a public telephone booth, what are
the probabili-es that you will have to wait
(a) more than 10 minutes?
(b) between 10 and 20 minutes?

Special distribu-ons: exponen-al
Mean and variance of exponen-al random variable:

Proof:
Using integra-on by parts, let






32
 Special distribu-ons: exponen-al

Memorylessness of exponen-al distribu-on:

Let X be an exponen-ally distributed random variable
with parameter . We have: for all s,t ≥0,

Remark: the geometric distribu-on is also memoryless

(in this case, s,t > 0 are integers).

 Special distribu-ons: exponen-al

Example: Consider a post office which is staffed by two

clerks. Suppose that when Mr. C enters the post
office, he discovers that Mr. A is being served by one
of the clerks and Mr. B by the other one. Suppose also
that C is told that his service will begin as soon as
either A or B leaves. If the amount of -me that a clerk
spends with a customer is exponen-ally distributed
with parameter , what is the probability that, of the
three customers, C is the last to leave the post office?
 Special distribu-ons: gamma

Gamma distribu-on:
Defini-on: A random variable X is said to be gamma
distributed with parameters (α,λ), denoted as
X~Gamma(α,λ), if its pdf is given by





Special distribu-ons: gamma
The gamma func-on Γ is given by
One important property of Γ is:
For integer values of α=n, we get Γ(n)=(n-1)!



 Special distribu-ons: gamma

Remark:
1) It is trivial that by the defini-on of Γ.

2) You are not required to memorize this pdf, or the

defini-on of the Gamma func-on. But once given the
density, you should be able to e.g. calculate the
moments.

Example: Compute the moments of X~Gamma(α,λ).

 Special distribu-ons: beta

Beta distribu-on
A r.v. is said to have a beta distribu-on if its density is
given by


where is called the beta
func-on, so that f(x) integrates to 1.
 Special distribu-ons: beta

Beta distribu-on

 Distribu-on of a func-on of a r.v.

4.3 Distribu-on of a func-on of a random variable

Suppose that we know the distribu-on of X, and we
want to find the distribu-on of g(X). To do so, it is
necessary to express the event that g(X)≤y (for each y)
in terms of X being in some set.
 Distribu-on of a func-on of a r.v.

Example: Suppose that a random variable X has pdf

Let Y=2X+1, what is the pdf of Y?
 Distribu-on of a func-on of a r.v.

Example: Suppose a con-nuous r.v. has a strictly

increasing cumula-ve distribu-on func-on F. What is
the distribu-on of F(X)?
(Note: “strictly increasing” is not necessary to find the
distribu-on of F(X), but it makes the argument
simpler)
Solu-on: