Probability and Statistics I

STAT 3600 – Fall 2021

Instructor:

Le Chen
[email protected]

Last updated on

October 18, 2021

Auburn University
Auburn AL

0
Chapter 3. Continuous Distributions

1
 Chapter 3. Continuous Distributions

§ 3.1 Random Variables of the Continuous Type

§ 3.2 The Exponential, Gamma, and Chi-Square Distributions

§ 3.3 The Normal Distributions

§ 3.4 Additional Models

2
 Chapter 3. Continuous Distributions

§ 3.1 Random Variables of the Continuous Type

§ 3.2 The Exponential, Gamma, and Chi-Square Distributions

§ 3.3 The Normal Distributions

§ 3.4 Additional Models

3
Definition 3.1-1 X is a continuous r.v. only if its range contains an interval (either
finite or infinite) of real numbers.
Examples of continuous r.v. are the length of time it takes when waiting in
line to buy frozen yogurt, the weight of a “1-pound” package of hot dogs,
etc.

4
Definition 3.1-2 The function f (x) is called the probability density function (pdf)
of the continuous r.v. X , with space S if it satisfies the following conditions
(a) fR(x) ≥ 0, ∀x ∈ S.
(b) S f (x)dx = 1.
(c) If (a, b) ⊂ S, then the probability of the event {a < X < b} is
Z b
P(a < X < b) = f (x)dx.
a

The corresponding distribution of probability is said to be of continuous type.

5
Definition 3.1-3 The cumulative distribution function (cdf) of a random variable X
of the continuous type, defined in terms of the X , is given by
Z x
F (x) = P(X ≤ x) = f (t)dt, −∞ < x < ∞.
−∞

Remark 3.1-1 By the fundamental theorem of calculus, we have, for x values for
which the derivative F 0 (x) exists,

F 0 (x) = f (x).

6
Example 3.1-1 Let X be a continuous r.v. with pdf

kx if 0 < x < 1


f (x) = 
0
 otherwise,

where k is a constant.
(a) Determine the value of k and sketch f (x).
(b) Find and sketch the corresponding cdf F (x).
(c) Find P (1/4 < X ≤ 2).

7
Remark 3.1-2 Note that if X is continuous r.v.

P(X = b) = 0, for all real values of b.

As a consequence,

P(a ≤ X ≤ b) = P(a < X < b)

= P(a ≤ X < b)
= P(a < X ≤ b)
= F (b) − F (a).

8
Example 3.1-2 The pdf of a continuous r.v. X is given by




1/3 if 0 < x < 1,
2/3 if 1 < x < 2,

f (x) = 




0
 otherwise.

Find the corresponding cdf F (x) and sketch f (x) and F (x).

9
Expected value of X , or the mean of X , is
Z ∞
µ = E(X ) = xf (x)dx.
−∞

The variance of X is
Z ∞
σ2 = Var (X ) = E[(X − µ)2 ] = (x − µ)2 f (x)dx,
−∞

and the standard deviation of X is

q
σ= Var(X ).

The moment-generating function, if it exists, is

Z ∞
M(t) = etx f (x)dx.
−∞

10
Example 3.1-3 Find the mean and variance of the r.v. X of Example 3.1-1.

11
 Uniform distribution

Definition 3.1-4 A r.v. X is called a uniform r.v. over (a, b) if its pdf is given by

 b−a if a < x < b

 1
f (x) = 
0
 otherwise.

12
 The corresponding cdf of uniform r.v. X is

if x ≤ a




0
if a < x < b

 x−a
F (x) = 
 b−a
if x ≥ b.


1


The mean and variance of the uniform r.v. X are

a+b (b − a)2
µ = E(X ) = and σ2 = Var (X ) = .
2 12

The moment-generating function is

 t(b−a) if t , 0
 tb ta

 e −e
M(t) = 
1
 if t = 0.

13
Example 3.1-4 Let f (x) = 1/2, −1 ≤ x ≤ 1, and 0 otherwise be the pdf of X .
(a) Graph the p.d.f. and distribution function.
(b) Find the mean, variance, and mgf.

14
Example 3.1-5 If the moment-generating function of X is

e5t − e4t
M(t) = , t ,0 and M(0) = 1.
t
Find (a) E(X ); (b) Var (X ); and (c) P (4.2 < X ≤ 4.7).

15
Example 3.1-6 Let X have the pdf

xe−x

 if 0 ≤ x < ∞,
f (x) = 
0
 otherwise.

Find the mgf, mean, and variance of r.v. X .

16
Definition 3.1-5 The (100p)th percentile is a number πp such that the area under
f (x) to the left of πp is p, that is,
Z πp
p= f (x)dx = F (πp ).
−∞

The 50th percentile is called the median. We let m = π0.50 . The 25th and 75th
percentiles are called the first and and q3 = π0.75 .

17
Example 3.1-7
Let X have the pdf

e−x−1

 −1 < x < ∞,
f (x) = 
0
 x ≤ −1.

(a) Find P (X ≥ 1).

(b) Find mgf.
(c) Find the mean and variance.
(d) Find the first quartile, the second or median, and the third quartile.

18
Exercises from textbook:section 3.1: 1, 2, 3, 4, 6, 7, 8, 9, 10, 16, 18, 20.

19
 Chapter 3. Continuous Distributions

§ 3.1 Random Variables of the Continuous Type

§ 3.2 The Exponential, Gamma, and Chi-Square Distributions

§ 3.3 The Normal Distributions

§ 3.4 Additional Models

20
 Chapter 3. Continuous Distributions

§ 3.1 Random Variables of the Continuous Type

§ 3.2 The Exponential, Gamma, and Chi-Square Distributions

§ 3.3 The Normal Distributions

§ 3.4 Additional Models

21
 Exponential distribution

Definition 3.2-1 A r.v. X is called an exponential r.v. with parameter λ(> 0) if its
pdf is given by

 θ e−x/θ , 0 ≤ x < ∞,

 1
f (x) = 
0
 0 > x.

22
The corresponding distribution function is

if −∞ < x < 0,

0


F (x) = 
1 − e−x/θ if 0 ≤ x < ∞.


and the moment-generating function is

1 1
M(t) = , ∀t < ,
1 − θt θ
from which, we see that

µ = M 0 (0) = θ and σ2 = M 00 (0) − [M 0 (0)]2 = θ2 .

23
Remark 3.2-1 It is useful to note that for an exponential random variable, X , we
have

P(X > x) = 1 − F (x) = 1 − (1 − e−x/θ ) = e−x/θ when x > 0.

Hence, W , the waiting time until the first occurrence in a Poisson process(λ > 0), has
an exponential distribution with parameter θ = 1/λ.

24
Example 3.2-1 It is known that the time (in hours) between consecutive traffic
accidents can be described by the exponential r.v. X with parameter θ = 60. Find
(a) P(X ≤ 60); (b) P(X > 120); (c) P(10 < X < 100); (d) the median time.

25
Example 3.2-2 Suppose that a certain type of electronic component has an
exponential distribution with a mean life of 500 hours. If X denotes the life of this
component (or the time to failure of this component), then
(a) What is the probability that this component will last at least 300 hours.
(b) Given that it has lasted at least 300 hours, what is the conditional probability that
it will last at least another 600 hours.

26
The most interesting property of the exponential distribution is its
”memoryless” property,

P(X > x + t|X > t) = P(X > x).

By this we mean that if the lifetime of an item is exponentially distributed,

then an item which has been in use for some hours is as good as a new item
with regard to the amount of time remaining until the item fails.

The exponential distribution is the only continuous distribution which

possesses this memoryless property.

27
 Gamma distribution

Definition 3.2-2 The gamma function is defined by

Z ∞
Γ(t) = y t−1 e−y dy , 0 < t.
0

Remark 3.2-2 For t > 1, by integration-by-parts, one finds the recursive relation:

Γ(t) = (t − 1)Γ(t − 1).

For example, Γ(5) = 4Γ(4), Γ(4) = 3Γ(3), Γ(3) = 2Γ(2), Γ(2) = 1Γ(1) and
Γ(1) = 1 Thus, when n is a positive integer, we have

Γ(n) = (n − 1)!

28
Here are some other special values:
√
4 π
!
3
Γ − =
2 3
!
1 √
Γ − = −2 π
2
!
1 √
Γ = π
2
Γ (1) = 0! = 1
√
π
!
3
Γ =
2 2
Γ (2) = 1! = 1
√
3 π
!
5
Γ =
2 4

and
1 1 1 1
= = = =0
Γ(−3) Γ(−2) Γ(−1) Γ(0)

29
Definition 3.2-3 A r.v. X has a gamma distribution if its pdf is defined by
1
f (x) = x α−1 e−x/θ , 0 ≤ x < ∞.
Γ(α)θα

Remark 3.2-3 W , the waiting time until the αth occurrence in a Poisson
process(λ > 0), has a gamma distribution with parameters α and θ = 1/λ.
The moment generating function of X is
1
M(t) = , t < 1/θ
(1 − θt)α
The mean and the variance are

µ = αθ and σ2 = αθ2 .

30
Example 3.2-3 Telephone calls enter a college switchboard at a mean rate of
two-thirds of a call per minute according to a Poisson process. Let X denote the
waiting time until the tenth call arrives.
(a) What is the pdf of X ?
(b) What are the moment-generating function, mean, and variance of X .

31
Example 3.2-4 If X has a gamma distribution with θ = 4 and α = 2, find
P(X < 5).

32
Example 3.2-5 In a medical experiment, a rat has been exposed to some radiation.
The experimenters believe that the rat’s survival time X (in weeks) has the pdf

x2
f (x) = e−x/120 0 < x < ∞.
2 × 1203
157 −5/6
(a) What is the probability that the rat survives at least 100 weeks? (Ans. 72 e )
(b) Find the expected value of the survival time. (Ans: 360)

33
 Chi-square distribution

Definition 3.2-4 A r.v. X has a chi-square distribution, denoted as X ∼ χ2 (r ), if its

pdf is defined by
1
f (x) = x r /2−1 e−x/2 , 0 ≤ x < ∞.
Γ(r /2)2r /2

Remark 3.2-4 The chi-square distributions is a gamma distribution with θ = 2 and

α = r /2.

The mean and the variance of this chi-square distribution are, respectively,
r  r 
µ = αθ = 2 = r and σ2 = αθ2 = 22 = 2r
2 2
and the moment-generating function is

M(t) = (1 − 2t)−r /2 ∀t < 1/2.

34
Example 3.2-6 If X ∼ χ2 (23), find the following:
(a) P(14.85 < X < 32.01). ( Use Table IV)
(b) Constants a and b such that P(a < X < b) = 0.95 and P(X < a) = 0.025.
(c) The mean and the variance of X .
(d) χ20.05 (23) and χ20.95 (23).

35
Exercises from textbook: Section 3.2: 1, 3, 4, 5, 7, 8, 9, 10, 11, 13, 14, 15,
17, 18, 21, 22

36
37
 Chapter 3. Continuous Distributions

§ 3.1 Random Variables of the Continuous Type

§ 3.2 The Exponential, Gamma, and Chi-Square Distributions

§ 3.3 The Normal Distributions

§ 3.4 Additional Models

38
 Chapter 3. Continuous Distributions

§ 3.1 Random Variables of the Continuous Type

§ 3.2 The Exponential, Gamma, and Chi-Square Distributions

§ 3.3 The Normal Distributions

§ 3.4 Additional Models

39
Definition 3.3-1 A r.v. X follows the normal distribution with parameters µ ∈ R
and σ > 0, denoted as X ∼ N(µ, σ2 ), if its pdf is defined by

(x − µ)2
" #
1
f (x) = √ exp − , −∞ < x < ∞.
σ 2π 2σ2

40
41
 The moment-generating function of X is

σ2 t 2
!
M(t) = exp µt + .
2

Consequently,

E(X ) = M 0 (0) = µ,

Var(X ) = M 00 (0) − [M 0 (0)]2 = µ2 + σ2 − µ2 = σ2 .

That is, the parameters µ and σ2 in the pdf of X are

the mean and the variance of X .

42
 Standard normal distribution
If Z ∼ N(0, 1), we say that Z has a standard normal distribution.

The distribution function of Z is

Z z
1 2
Φ(z) = P(Z ≤ z) = √ e−w /2 dw.
−∞ 2π
Values of Φ(z) for z ≥ 0 is given in Table Va in the appendix of the
textbook.

Because of the symmetry of the standard normal pdf it is true that

Φ(−z) = 1 − Φ(z), ∀z ∈ R.

Again, because of symmetry of the standard normal pdf, when z > 0,

Φ(−z) = P(Z ≤ −z) = P(Z > z)

can be read directly from Table Vb.

43
Example 3.3-1 If Z ∼ N(0, 1), find
(a) P(0.53 < Z ≤ 2.06).
(b) P(−0.79 ≤ Z < 1.52).
(c) P(Z > −1.77).
(d) P(Z > 2.89).

44
In statistical applications, we are often interested in finding a number zα
such that

P(Z ≥ zα ) = α,

where Z is N(0, 1) and α is usually less that 0.5. That is, zα is the
100(1 − α)th percentile (sometimes called the upper 100α percent point) for
standard normal distribution.

45
Example 3.3-2 Find the values of (a) z0.01 ; (b) −z0.005 ; (c) z0.0475 .

46
 X −µ
Theorem 3.3-1 If X ∼ N µ, σ2 , then Z =
 
∼ N(0, 1).
σ

47
Theorem 3.3-1 can be used to find probabilities relating to X ∼ N µ, σ2 , as
 

follows:
a−µ X −µ b−µ b−µ a−µ
! ! !
P(a ≤ X ≤ b) = P ≤ ≤ =Φ −Φ
σ σ σ σ σ

since (X − µ)/σ is N(0, 1).

48
Example 3.3-3 If the moment-generating function of X is
M(t) = exp(166t + 200t 2 ), find (a) The mean of X ; (b) The variance of X ; (c)
P(170 < X < 200); (d) P(148 < X < 172); (e) P(|X − 166| < 40).

49
Theorem 3.3-2 If X ∼ N(µ, σ2 ), then V = (X − µ)2 /σ2 = Z 2 ∼ χ2 (1).

50
Example 3.3-4 If Z is N(0, 1), find values of c such that (a) P(Z ≥ c) = 0.025;
(b) P(|Z | ≤ c) = 0.9.

51
Exercises from textbook: Section 3.3: 1, 3, 5cd, 6, 7, 8, 9, 11, 12, 13, 17.

52
53
54
 Chapter 3. Continuous Distributions

§ 3.1 Random Variables of the Continuous Type

§ 3.2 The Exponential, Gamma, and Chi-Square Distributions

§ 3.3 The Normal Distributions

§ 3.4 Additional Models

55
 Chapter 3. Continuous Distributions

§ 3.1 Random Variables of the Continuous Type

§ 3.2 The Exponential, Gamma, and Chi-Square Distributions

§ 3.3 The Normal Distributions

§ 3.4 Additional Models

56
Let λ(x) ≥ 0, then the following function is a legal pdf:
Z w !
g(w) = λ(w) exp − λ(t)dt , w ≥ 0
0

λ(x) Distribution
1/θ Exponential distribution with parameter θ.
αw α−1
Weibull distribution
βα
aebw Gompertz distribution

57
 Weibull distribution

Definition 3.4-1 A random variable X follows the Weibull distribution with shape
parameter α > 1 and scale parameter θ > 0 if its density (pdf) function is
!α !
αx α−1 x
fX (x) = exp − , x > 0,
βα β
and 0 otherwise.

58
Theorem 3.4-1 The mean and variance for the Weibull distribution are
!  ! ( !)2 
1 2 1
µ = βΓ 1 + and σ2 = β2 Γ 1 +
 
− Γ 1+
α α α


59
Example 3.4-1 In a medical experiment, a rat has been exposed to some radiation.
The experimenters believe that the rat’s survival time X (in weeks) has the pdf

3x 2 −(x/120)3
f (x) = e 0 < x < ∞.
1203
(a) What is the probability that the rat survives at least 100 weeks? (Ans. e−(125/216) )
(b) Find the expected value of the survival time. (Ans: 120Γ(4/3))

60
 Gompertz Law

Definition 3.4-2 A random variable X follows the Gompertz distribution with

parameters (a, b), a > 0 and b > 0, if its density (pdf) function is
 a a
fX (x) = aebx exp − ebx + , x > 0,
b b
and 0 otherwise.

61