Recall that a random variable is a function X : R which
associates each outcome with a real number.
Since the outcome of the experiment with sample space is random,
the number produced by the function is random as well.
You ought to be familiar with the following classification of random
variables:
RANDOM VARIABLES
A discrete random variable can take on only a finite or at most a
countably infinite number of values.
A continuous random variable can take on a continuum of values.
AY 20142015 (Semester 2)
2 RANDOM VARIABLES
1 / 21
AY 20142015 (Semester 2)
2.1
O UTLINE
2.1
RANDOM VARIABLES
2 / 21
Discrete Random Variables
Let X be a discrete random variable taking values x1 , x2 , . . . .
The probability properties of X are completely described by the
probability mass function
P (or frequency function) p such that
p(xi ) = P(X = xi ) and i p(xi ) = 1.
Discrete Random Variables
Bernoulli Random Variables
2.2
Continuous Random Variables
A Bernoulli random variable takes on only two values, 0 and 1, with
probabilities 1 p and p, respectively.
2.3
Functions of a Random Variable
Its frequency function is thus
(
px (1 p)1x ,
p(x) =
0,
AY 20142015 (Semester 2)
2 RANDOM VARIABLES
3 / 21
AY 20142015 (Semester 2)
if x = 0 or x = 1,
otherwise.
RANDOM VARIABLES
4 / 21
2.1
Discrete Random Variables
2.1
Discrete Random Variables
Binomial Distribution
Suppose that n independent experiments (or trials) are performed,
where n is a fixed number, and that each trial results in a success
with probability p and a failure with probability 1 p.
The total number of successes, X , is a binomial random variable
with parameters n and p.
The probability that X = k is given by
n k
p(k) =
p (1 p)nk ,
k
k = 0, 1, . . . , n.
If X1 , . . . , Xn are independent Bernoulli random variables with
P(Xi = 1) = p, then Y = X1 + + Xn is a binomial random variable.
AY 20142015 (Semester 2)
2.1
2 RANDOM VARIABLES
5 / 21
AY 20142015 (Semester 2)
2.1
Discrete Random Variables
2 RANDOM VARIABLES
6 / 21
Discrete Random Variables
Negative Binomial Distribution
Geometric Distribution
Suppose that a sequence of independent Bernoulli trials is performed
and X is the total number of trials up to and including the first
success.
Then X is a geometric random variable with
p(k) = P(X = k) = p(1 p)k1 ,
AY 20142015 (Semester 2)
Figure 2.3: Binomial frequency functions, (a) n = 10 and p = 0.1 and
(b) n = 10 and p = 0.5.
2 RANDOM VARIABLES
This is a generalization of the geometric distribution.
Suppose that a sequence of independent trials each with probability of
success p is performed until there are r successes in all.
Let X denote the total number of trials. Then
k 1 r
P(X = k) =
p (1 p)kr ,
k = r , r + 1, . . . .
r 1
k = 1, 2, 3, . . . .
7 / 21
AY 20142015 (Semester 2)
RANDOM VARIABLES
8 / 21
2.1
Discrete Random Variables
2.1
Discrete Random Variables
Hypergeometric Distribution
Suppose that an urn contains n balls, of which r are black and n r
are white.
Let X denote the number of black balls drawn when taking m balls
without replacement.
Figure 2.4: The probability mass
function of a geometric random
variable with p = 1/9.
AY 20142015 (Semester 2)
2.1
X is a hypergeometric random variable with parameters r , n, and m,
and
r
nr
k
mk
P(X = k ) =
.
n
m
Figure 2.5: The probability mass
function of a negative binomial
random variable with p = 1/9 and
r = 2.
2 RANDOM VARIABLES
Recall Example 1.4.2I (Capture/Recapture Method).
9 / 21
Discrete Random Variables
2.1
Poisson Distribution
RANDOM VARIABLES
10 / 21
Discrete Random Variables
The Poisson distribution has been used in many areas, such as
A Poisson random variable with parameter > 0 has the frequency
function
k e
P(X = k ) =
,
k = 0, 1, 2, . . . .
k!
The Poisson frequency function can be used to approximate binomial
probabilities for large n and small p.
AY 20142015 (Semester 2)
AY 20142015 (Semester 2)
2 RANDOM VARIABLES
11 / 21
analysis of telephone systems,
modeling the number of alpha particles emitted from a radioactive
source during a given period of time, and
modeling the number of freak accidents for a large population of
people in a given time period.
AY 20142015 (Semester 2)
RANDOM VARIABLES
12 / 21
2.2 Continuous Random Variables
2.2
For a continuous random variable, the role of the frequency function is
taken by a density
R function, f (x), such that f (x) 0, f is piecewise
continuous, and f (x) dx = 1.
P(a < X < b) =
Exponential Distribution
The exponential density function is
(
ex ,
f (x) =
0,
If X is a random variable with a density function f , then
Z
Continuous Random Variables
f (x) dx.
x 0,
x < 0.
The cumulative distribution function (cdf) of a continuous random
variable X is defined by
Z x
F (x) = P(X x) =
f (u) du.
The exponential distribution is often used to model lifetimes or
waiting times and the memoryless property of the distribution has
important implications here:
P(T > t + s | T > s) =
e(t+s)
= et = P(T > t).
es
From the fundamental theorem of calculus, if f is continuous at x,
f (x) = F 0 (x).
AY 20142015 (Semester 2)
2 RANDOM VARIABLES
13 / 21
2.2 Continuous Random Variables
AY 20142015 (Semester 2)
2.2
RANDOM VARIABLES
14 / 21
Continuous Random Variables
Normal Distribution
Gamma Distribution
The normal distribution plays a central role in probability and statistics.
The central limit theorem (Ch 5) justifies the use of the normal
distribution in many applications.
The gamma density function depends on two parameters, > 0
(shape) and > 0 (scale):
x 1 ex ,
x 0,
f (x) = ()
0,
x < 0,
in which (t) =
R
0
u t1 eu
The density function of the normal distribution depends on two
parameters, R (the mean) and > 0 (the standard deviation):
1
2
2
f (x) = e(x) /2 ,
2
du, t > 0, is the gamma function.
< x < .
If = 1, the gamma density coincides with the exponential density.
The cdf cannot be evaluated in closed form from this density function
and must be found numerically.
Gamma densities provide a fairly flexible class for modeling
nonnegative random variables.
The special case for which = 0 and = 1 is called the standard
normal distribution. We denote its cdf by (which is tabulated) and
its density by .
AY 20142015 (Semester 2)
2 RANDOM VARIABLES
15 / 21
AY 20142015 (Semester 2)
RANDOM VARIABLES
16 / 21
2.2 Continuous Random Variables
2.2
Continuous Random Variables
Beta Distribution
The beta density is useful for modeling random variables that are
restricted to the interval [0, 1]:
f (u) =
(a + b) a1
u
(1 u)b1 ,
(a)(b)
0 u 1.
Note that the case a = b = 1 is the uniform distribution.
The beta distribution is important in Bayesian statistics (see
Example 3.5.2E).
Figure 2.18: Beta density functions for various values of a and b: (a) a = 2,
b = 2; (b) a = 6, b = 2; (c) a = 6, b = 6; and (d) a = 0.5, b = 4.
AY 20142015 (Semester 2)
2.3
2 RANDOM VARIABLES
17 / 21
AY 20142015 (Semester 2)
RANDOM VARIABLES
18 / 21
2.3 Functions of a Random Variable
Functions of a Random Variable
Proposition A
If X N(, 2 ) and Y = aX + b, then Y N(a + b, a2 2 ).
Since
It follows that Z =
N(0, 1) and FX (x) =
P(120 < X < 130) = P
x
.
120 100
X 100
130 100
<
<
15
15
15
= P(1.33 < Z < 2)
Example A
= (2) (1.33)
Scores on a certain standardized test, IQ scores, are approximately
normally distributed with mean = 100 and standard deviation
= 15.
= 0.9772 0.9082 = 0.069,
approximately 7% of the population will have scores in this range.
An individual is selected at random. What is the probability that his
score X satisfies 120 < X < 130?
AY 20142015 (Semester 2)
2 RANDOM VARIABLES
19 / 21
AY 20142015 (Semester 2)
RANDOM VARIABLES
20 / 21
2.3
Functions of a Random Variable
Propositions C & D
1
Let Z = F (X ). Then Z has a uniform distribution on [0, 1].
Let U be uniform on [0, 1], and let X = F 1 (U). Then the cdf of X
is F .
The results follow directly from
P(Z z) = P(F (X ) z) = P(X F 1 (z)) = F (F 1 (z)) = z,
which is the uniform cdf, and
P(X x) = P(F 1 (U) x) = P(U F (x)) = F (x).
To generate random variables with cdf F , the second result suggests
that we just apply F 1 to uniform random numbers. This is quite
practical as long as F 1 can be calculated easily.
AY 20142015 (Semester 2)
2 RANDOM VARIABLES
21 / 21