Chap0 Contineous

Table of contents
Continuous Random Variable
anis.rezguii@gmail.com
INSAT
Mathematics Department
September 20, 2022
anis.rezguii@gmail.com Continuous Random Variable

Expected learning outcomes
Definitions
The Normal distribution
The Exponential distribution
The Gamma distribution
The ”Chisquare” distribution
The Td distribution of Student
The Fisher distribution
Table of contents
1 Expected learning outcomes
2 Definitions
The probability density function
The cumulative distribution function
The mathematical expectation
The variance and the standard deviation
3 The Normal distribution
4 The Exponential distribution
5 The Gamma distribution
6 The ”Chisquare” distribution
7 The Td distribution of Student
8 The Fisher distribution
Definitions
After this lecture the student should:
Understand the concept of continuous random variable and

the difference between them and the discrete ones.
Know most usual examples of continuous random variable.
be able to chose which model is adequate for a given situation.

Definitions
The Normal distribution The probability density function
The Exponential distribution The cumulative distribution function
The Gamma distribution The mathematical expectation
The ”Chisquare” distribution The variance and the standard deviation
Continuous random variable

Definition
Let (Ω, P) be a probability space and X be a mapping,
X : Ω −→ R. X is a continuous random variable if there exists a
positive piecewisea continuous function fX such that:
Z +∞
1 fX (x)dx = 1,
−∞
2 for any −∞ ≤ a ≤ b ≤ +∞ we have
Z b
P{X ∈ [a, b]} = fX (x)dx.
a
The function fX is called the Probability Density Function (PDF).

Definitions
Note
If X is a continuous random variable with a PDF fX we

denote
X ,→ fX (x)dx.
We also say that X follows the distribution characterized by

its PDF fX .

Definitions
Example: The uniform distribution

Let a < b two real numbers and, X a continuous random variable
is following the uniform distribution if its PDF is given by:

1

on [a, b] 1
fX = b−a = χ (x).
 0 if not. b − a [a,b]
It is easy to check that fX is a positive piecewise continuous

function and that it satisfies
Z Z b
1
fX (x)dx = dx = 1.
R b−a a
We denote:
X ,→ U([a, b]).

Definitions
Note
By definition of a continuous random variable, X , it is clear that

for any real number a ∈ R:
Z a
P{X = a} = fX (x)dx = 0.
a
We say that a continuous random variable doesn’t charge points,

however a discrete random variable charges points !

Definitions
Cumulative distribution function
Definition
Let X be a continuous random variable with a PDF given by fX .
We associate to X its Cumulative Distribution Function (CDF) FX ,
defined by:
FX : R −→ [0, 1] Z x
x 7−→ P{X ≤ x} = fX (t)dt.
−∞

Definitions
The example of the uniform distribution

Suppose X following the uniform distribution, X ,→ U([a, b]), its
CDF is defined by:
Z x
FU ([a,b]) (x) = χ[a,b] (t)dt.
−∞

Definitions

Suppose X following the uniform distribution, X ,→ U([a, b]), its
CDF is defined by:
Z x
FU ([a,b]) (x) = χ[a,b] (t)dt.
−∞
Check that:

 0x − a
 if x < a
FU ([a,b]) (x) = if x ∈ [a, b]
 b−a
if x > b.

1

Definitions
Note
1 The CDF, FX , of a given continuous random variable X
satisfies:
i. It is non-decreasing.
ii. lim FX (x) = 1 and lim FX (x) = 0.
x→+∞ x→−∞
2 The CDF of a given continuous random variable, X , is
differentiable wherever the associated PDF is, and we have:
dFX (x)
= fX (x).
dx
3 The latter relation shows that the CDF determines the PDF,
so, we can say that the CDF and the PDF are simply
equivalent.
Definitions
Exercise
Let X ,→ ke −|x| dx be a continuous random variable.

1 Evaluate k.
2 Determine its cumulative distribution function, FX .
3 Deduce the probability density function, fY , of Y = X 2 .

Definitions
The mathematical expectation
Definition
Let X be a continuous random variable X ,→ fX (x)dx, its
mathematical expectation is the quantity:
Z
E(X ) = xfX (x) dx.
R

Definitions
Notes
1 If any random experiment is repeated many times we expect,
in the ”usual” cases, that most of X’s values are ”close” to
the mathematical expectation E(X ). Thus the mathematical
expectation of a d.r.v is considered as a parameter of central
tendency.
2 If X = c is constant, then its expectation E(X ) = c.
3 E(λX + c) = λE(X ) + c for any real numbers λ and c.
4 Let g : R −→ R be a continuous function, then the
mathematical expectation of the new random variable
Y = g (X ) is given by:
Z −∞
E(g (X )) = g (x)fX (x)dx.
+∞

Definitions
Examples
1 Let X ,→ U([a, b]) then its expectation is:
a+b
E(X ) = .
2
1 −|x|
2 Let X ,→ e dx, then its expectation is:
2
E(X ) = 0.

Definitions
The Markov Inequality

We have exactly the same result as for discrete random variables.
The Markov inequality gives a rough estimation of some
probabilities by the only knowledge of the expectation.
Theorem
Let (Ω, Σ, P) be a probability space, X a positive d.r.v and λ a
positive real number. Then
E(X )
P{X ≥ λ} ≤ .
λ

Definitions
The Markov Inequality

We have exactly the same result as for discrete random variables.
The Markov inequality gives a rough estimation of some
probabilities by the only knowledge of the expectation.
Theorem
Let (Ω, Σ, P) be a probability space, X a positive d.r.v and λ a
positive real number. Then
E(X )
P{X ≥ λ} ≤ .
λ
Note: The only knowledge of the expectation doesn’t determine,

in general, the probability distribution of a continuous random
variable. anis.rezguii@gmail.com Continuous Random Variable
Definitions
The variance
Definition
Let X be a c.r.v with as a PDF fX (x). Its variance, denoted by
V(X ) or σX2 , is the following non negative quantity:
Z
V(X ) = σX2 = (x − E(X ))2 fX (x)dx.
R
Its standard deviation is
p
σX = V(X ).

Definitions
Notes
1 We expect, after many repetitions of the experiment, and in
the ”usual” cases, that most of X’s values lies in the interval
[E(X ) − σX , E(X ) + σX ].
2 The variance and the standard deviation are considered to be

parameters of dispersion which means that they give idea
about how the X ’s values, in case of repetitions, are spread
out around the expectation.
3 If the variance is finite we have the following new formulation:
V(X ) = E(X 2 ) − E(X )2 .

Definitions
The Tchebychev Inequality
Theorem
Let (Ω, Σ, P) be a probability space, and X a c.r.v, then for any
λ > 0 we have
V(X )
P{|X − E (X )| ≥ λ} ≤ .
λ2

Definitions
Note
Tchebychev inequality is often stated as follows:
1
P{|X − E (X )| ≥ nσ} ≤
n2
which can be read
1
P {X ∈ [E(X ) − nσ, E(X ) + nσ]} > 1 −
n2
If for example we take n = 6 we obtain that for any random
variable X , which means for any phenomenon we have the
following estimation:
1
P {X ∈ [E(X ) − 6σ, E(X ) + 6σ]} > 1 − = 97.22%
n2
.
Definitions
Let X ,→ U([a, b]) then:
V(X ) = E (X 2 ) − E (X )2
Z b 2
a+b
= x 2 dx −
a 2
(b − a)2
=
12

Definitions
The normal distribution
Definition
Let µ ∈ R, σ ∈ R∗+ and X a continuous random variable. We say
that X follows a normal distribution of parameters µ and σ and we
denote X ,→ N(µ, σ), if its probability density function is given by:
1 (x−µ)2
fµ,σ (x) = √ e − 2σ2 .
2πσ
If µ = 0 and σ = 1, X is said to follow the standard normal

distribution N(0, 1).

Definitions
Shape of its p.d.f curve

Definitions
Expectation and Variance

Suppose X ,→ N(0, 1) then:
Z −∞
1 x2
E(X ) = √ xe − 2 dx = 0
2π +∞
and
V(X ) = E(X 2 )
Z
1 x2
= √ x 2 e − 2 dx
2π R
Z +∞
−2 x2
= √ xd(e − 2 )
2π 0
2 +∞
Z +∞
−2 − x2 2 x2
= √ xe +√ e − 2 dx
2π 0 2π 0
= anis.rezguii@gmail.com
1. Continuous Random Variable
Definitions
The cumulative distribution function

Suppose Z following the standard normal distribution,
Z ,→ N(0, 1), then its CDF is given by:
Z x
t 2 dt
ϕ(x) = P{Z ≤ x} = P {Z ∈] − ∞, x]} = e− 2 √
−∞ 2π
It is easy to check that:
ϕ(x) + ϕ(−x) = 1.
ϕ has no explicit formula, to get its values we usually use

value’s table.
Definitions
0123456
789
28

8
106

45 6

anis.rezguii@gmail.com

Continuous

Random
Variable

Definitions
Use of the value’s table
1 Evaluate ϕ(x) = P{Z ≤ x} for a given x ∈ R:

ϕ(0) = 0.5
ϕ(1.64) = 0.9495
2 Find out x such that ϕ(x) = p for a given percentage p:
p = 95%, x = 1.64
p = 99%, x = 2.33
3 Evaluate P{−x ≤ z ≤ x} for a given x ∈ R:
Note that P{−x ≤ z ≤ x} = ϕ(x) − ϕ(−x) = 2ϕ(x) − 1
For x = 1.64, P{−1.64 ≤ z ≤ 1.64} = 0.899
For x = 2.5, P{−2.5 ≤ z ≤ 2.5} = 0.9876

Definitions
Linear transformations
The normal distribution is stable under linear transformation:

If Z ,→ N(µ, σ) then
Z −µ
,→ N(0, 1)
σ
If Z ,→ N(0, 1) then
σ × Z + µ ,→ N(µ, σ).

Definitions
The case of non-standard normal distribution
Suppose that Z ,→ N(10, 2).

1 Evaluate P{Z ≤ x}:
P{Z ≤ 4} = P{ Z −10
2 ≤ 4−10
2 } = ϕ(−3) = 0.13%
P{Z ≤ 11} = P{ Z −11

2 ≤ 11−10
2 } = ϕ(0.5) = 69.15%
2 Find out x such that P{Z ≤ x} = p for a given percentage p:

p = 95%, x = 13.289
p = 99%, x = 14.652

Definitions
Example: Quality Control

We control two dimensions of pieces produced in a factory. The
two dimensions x and y should measure respectively 650 mm and
830 mm. We tolerate on each dimension an error of ±0.1mm.
Measurements of a sample of pieces give the records below:
dimension mean standard deviation
x 650.01mm 0.05mm
y 830.02mm 0.06mm
Suppose that the theoretical models of the two dimensions are
both normally distributed.
1 Evaluate the theoretical percentage such that the dimension x
is acceptable.
2 Do the same thing for the dimensiony .
3 Evaluate the theoretical percentage of the defect pieces.

Definitions
Solution
1. The dimension x is acceptable if x ∈ [649.9 ; 650.1]. The
theoretical distribution of x is normal N(650, 01 ; 0.05). Then

649.9 − 650.01 x − x̄
P{649.9 ≤ x ≤ 650.1} = P ≤
0.05 σx

650.1 − 650.01
≤
0.05

x − x̄
= P −2.2 ≤ ≤ 1.8
σx
x − x̄
or ,→ N(0, 1) so
σx
P{649.9 ≤ x ≤ 650.1} = ϕ(1, 8) − ϕ(−2, 2)
= ϕ(1, 8) + ϕ(2, 2) − 1 = 95%
Definitions
Still solution
2. We do the same for the dimension y :
P{829.9 ≤ y ≤ 830.1} = 88%.
3. A piece is good if both dimensions x and y are acceptable, so

we get:
P{649.9 ≤ x ≤ 650.1 , 829.9 ≤ y ≤ 830.1} = 95%×88% = 84%.

The percentage of defect pieces is then:
100% − 84% = 16%.

Definitions
Exponential distribution
Definition
Let X be a continuous random variable and λ ∈ R∗+ a given
parameter. X is said to follow an exponential distribution if its
probability density function is given by:
λe −λx if x ∈ R+

fX =
0 otherwise
we denote X ,→ E(λ).

Definitions
The PDF curve
Exponential PDF curve

0.5
0.45
0.4
λ = 0.5
0.35
λ = 0.33
0.3
0.25
0.2
0.15
0.1
0.05
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Definitions
The exponential cumulative distribution function
Let X ,→ E(λ), it is easy to check that the cumulative distribution

function of X is given by:
1 − e −λx

if x ∈ R+
FX =
0 elsewhere

Definitions
Expectation and variance
Let X ,→ E(λ) then

1
1
E(X ) = .
λ
2
1
V(X ) = .
λ2

Definitions

Definition
A continuous random variable X follows a Gamma distribution of
parameter r > 0 if its density function has the form
1 r −1 −x
fX (x) = x e ,
Γ(r )
where the function Gamma is defined, for any r > 0, by

Z +∞
Γ(r ) = e −x x r −1 dx.
0

Definitions
Exercise
1 Check that for any r > 0, Γ(r + 1) = r Γ(r ).
2 Deduce that Γ(n + 1) = n!.
3 Prove that
lim
+
Γ(r ) = +∞.
0
lim Γ(r ) = +∞.
+∞

Definitions
The parameters of the Gamma distribution

Expectation
Z +∞
1
E(X ) = x r e −x dx
Γ(r ) 0
Γ(r + 1)
= =r
Γ(r )
Variance
V(X ) = E(X 2 ) − E(X )2
Z +∞
1
= x r +1 e −x dx − r 2
Γ(r ) 0
Γ(r + 2)
= − r 2 = r (r + 1) − r 2 = r
Γ(r )
Definitions
The Chisquare distribution
Definition
Let U1 , U2 , · · · , Ud , d independent variables that follow all the
standard normal distribution N(0, 1). We call the ”Chisquare”
distribution with d degrees of freedom and denote χ2d the
distribution of the sum:
U12 + U22 + · · · + Ud2 .

Definitions
The CDF of the chisquare distribution has no explicit

formulae, to get its values we use either a statistical software
or a values table.
One can prove that
E(χ2d ) = d, V(χ2d ) = 2d
The PDF of the chisquare distribution is given by:

1 −x 1
gd (x) = d
exp x d− 2 1R (x).
d/2
2 Γ( 2 ) 2 +

Definitions
Case d=1
Let U ,→ N(0, 1) and x ≥ 0, the CDF of U 2 is given by
FU 2 (x) = P{U 2 ≤ x}
√ √
= P{− x ≤ U ≤ x}
Z √x −t 2 /2 Z √x −t 2 /2
e e
= √
√ t=2 √ dt
− x 2π 0 2π
this leads to
√ 2
∂FU 2 e − x /2 1 1 1
g1 (x) = (x) = 2 √ √ = √ e −x/2 √ 1R (x)
∂x 2π 2 x 2π x +

Definitions
Case d=2
Let U1 and U2 two independent r.v such that U1 , U2 ,→ N(0, 1),
then
FU 2 +U 2 (x) = P{U12 + U22 ≤ x}
1 2
ZZ
1 1 2 2
= exp − (u1 + u2 ) du1 du2
2π {u12 +u22 ≤x} 2
Z √xZ 2π Z √x
1 2
− r2 r2
= re drdθ = re − 2 dr
2π 0 0 0
Z √x 2
r
= −d e − 2 = 1 − e x/2 1R (x)
+
0
it follows that χ22 = E(1/2).

Definitions
Let Y and U two independent random variables that follow

respectively χ2d and the standard normal distribution N(0, 1). We
define the Student random variable with d degrees of freedom, Td
by:
U
q .
Y
d
Its probability density function is given by:

− d+1
1 Γ( d+1 x2

2 )
2
fd (x) = √ d
1+
dπ Γ( 2 ) d

Definitions
It is easy to show that for d > 1
E(Td ) = 0
and that for d > 2

d
V(Td ) = .
d −2

Definitions
T -distribution

Definitions
Exercise
1 Suppose d = 1 show that T1 follows the Cauchy distribution
given by:
1 1
f1 (x) =
π 1 + x2
2 Deduce that for d = 1 the Student distribution has no

expectation and has an infinite variance.
3 Suppose now d = 2, prove the expression of the density
function:
− 32
x2

1
f2 (x) = √ 1+
2 2 2

Definitions
The F-distribution
Definition
Let A and B two independent random variable that follows
respectively χ2d1 and χ2d2 for d1 and d2 two given integers. Then
the random variable
A/d1
F =
B/d2
defines the so-called Fisher distribution of degrees of freedom d1
and d2 . We denote
A/d1
= F ∼ F(d1 , d2 ).
B/d2

Definitions
If F ∼ F(d1 , d2 ) then its statistical parameters are given by

d1
E(F ) = for d2 > 2
d2 − 2
2d22 (d1 + d2 − 2)
V(F ) =
d1 (d2 − 2)2 (d2 − 4)

Definitions
The F-distribution

Chap0 Contineous

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chap0 Contineous

Uploaded by

Copyright:

Available Formats

Table of contents

Continuous Random Variable

September 20, 2022

anis.rezguii@gmail.com Continuous Random Variable

Expected learning outcomes

After this lecture the student should:

Understand the concept of continuous random variable and

Know most usual examples of continuous random variable.

be able to chose which model is adequate for a given situation.

anis.rezguii@gmail.com Continuous Random Variable

Continuous random variable

The function fX is called the Probability Density Function (PDF).

If X is a continuous random variable with a PDF fX we

We also say that X follows the distribution characterized by

anis.rezguii@gmail.com Continuous Random Variable

Example: The uniform distribution

It is easy to check that fX is a positive piecewise continuous

anis.rezguii@gmail.com Continuous Random Variable

By definition of a continuous random variable, X , it is clear that

We say that a continuous random variable doesn’t charge points,

anis.rezguii@gmail.com Continuous Random Variable

Cumulative distribution function

anis.rezguii@gmail.com Continuous Random Variable

The example of the uniform distribution

anis.rezguii@gmail.com Continuous Random Variable

The example of the uniform distribution

anis.rezguii@gmail.com Continuous Random Variable

Let X ,→ ke −|x| dx be a continuous random variable.

anis.rezguii@gmail.com Continuous Random Variable

The mathematical expectation

anis.rezguii@gmail.com Continuous Random Variable

anis.rezguii@gmail.com Continuous Random Variable

1 Let X ,→ U([a, b]) then its expectation is:

anis.rezguii@gmail.com Continuous Random Variable

The Markov Inequality

anis.rezguii@gmail.com Continuous Random Variable

The Markov Inequality

Note: The only knowledge of the expectation doesn’t determine,

anis.rezguii@gmail.com Continuous Random Variable

2 The variance and the standard deviation are considered to be

3 If the variance is finite we have the following new formulation:

V(X ) = E(X 2 ) − E(X )2 .

anis.rezguii@gmail.com Continuous Random Variable

The Tchebychev Inequality

anis.rezguii@gmail.com Continuous Random Variable

The example of the uniform distribution

Let X ,→ U([a, b]) then:

anis.rezguii@gmail.com Continuous Random Variable

The normal distribution

If µ = 0 and σ = 1, X is said to follow the standard normal

anis.rezguii@gmail.com Continuous Random Variable

Shape of its p.d.f curve

anis.rezguii@gmail.com Continuous Random Variable

Expectation and Variance

The cumulative distribution function

It is easy to check that:

ϕ has no explicit formula, to get its values we usually use

Use of the value’s table

1 Evaluate ϕ(x) = P{Z ≤ x} for a given x ∈ R:

anis.rezguii@gmail.com Continuous Random Variable

The normal distribution is stable under linear transformation:

anis.rezguii@gmail.com Continuous Random Variable

The case of non-standard normal distribution