Probabilistic Methods in

Engineering
Lecture 16
Dr. Maha Ali
• Exponential Distribution
and reliability Application
• Gaussian( normal
Distribution)
• Functions of Random
Variables
• Joint PDF
CDF of Exponential distribution

x α e
to

CDE EW ca E x
Application of Exponential Distribution to UB CA MB
Reliability A
• Suppose a system fails when any of its n components fail. Let Xi denote failure
time of ith component, i = 1,...,n
• Let Y denote failure time of system.
• Find cdf of Y
Y ᵗB EX t
ufx t
USE
notdisjoint
P Y t P X t U U xn t3
1 P Xp B n'East AND
FE.PE et3 PEXn

Fy T
1
II P Xist 1 II 11 FEB
Application of Exponential Distribution to
Reliability cont.
• If each Xi has exponential cdf with mean µ,

Fy t 1
II 1
i
File
ti
Fi t 1 e

1_É
EEN p Failt I
Fy It 1 Fé 1 et 1 é n
exponential CDF themost the mean lifetimeis reducedby afackfy
Gaussian Random Variables (Normal Random
Variables)
• The most important continuous random variable — the Gaussian
random variable (also known as the normal random variable).
Cx m
G
12 e

is the mean (average value)

is the variance (spread)
a is standard deviation
where
Cry 62 are parameters of the distribution.
µ 2 or X N M62
We Gaussian to say that X is drawn from a
Gaussian distribution of parameter (µ, σ2 ).
Gaussian R.V. cont.

fix 0

µ location paramete
scaling paramete
around 68 3 µ 22 2 Mtd
9 5.44 µ 28 µ 28
Gaussian R.V. Cont

Lf fact 1

if is reduced ftp.i.me
6 is enlarged
forgo a
Gaussian R.V. Cont.
Plotting Gasussian code
• On a computer, plotting the Gaussian PDF can be done by calling the
function pdf(’norm’,x) in MATLAB, and stats.norm.pdf in Python.
Mean and Variance
2 ), then E[X] =
• If X ∼ Gaussian(µ, σ2
µ
, and Var[X] =. 82
Gaussian µ 62 N µ 62
EEX
If find
If
I d

let y X M
EW
ta
M
E Ex M
Mean and Variance cont.
2 ), then E[X] = µ, and Var[X] = σ 2 .
• If X ∼ Gaussian(µ, σ2

VAREN put é dg
E
let y XI
Standard Gaussian
We need to evaluate the probability P[a ≤ X ≤ b] of a Gaussian random variable X in
many practical situations. This involves the integration of the Gaussian PDF, i.e.,
determining the CDF. Unfortunately, there is no closed-form expression of P[a ≤ X ≤
b] in terms of (µ, σ2 ). This leads to what we call the standard Gaussian.
The standard Gaussian (or standard normal) random variable X
has a PDF:
µ 0 22 1

IN
f e

Fx Ix f x dx P X n
CDF of Standard Gaussian
• The CDF of the standard Gaussian is defined as the Φ(·) function

é
x Fx IN IF I It
CDF of an arbitrary Gaussian
• With the standard Gaussian CDF, we can define the CDF of an
arbitrary Gaussian.
Let X ∼ N (µ, σ2 ). Then
Fix Q T
F IN P a I d
7
let y 5 a
1 ftp e dy

oct 1
CDF of an arbitrary Gaussian cont.

P a b P X b P X a

b
Fx Fala
b 919
The complementary CDF of the N(0, 1)
distribution is traditionally denoted by the letter
Q.
9 e dx

OF
In e dx 1
A x
a

idiemetters.tt
ii IC M 1 ow
Normality is preserved by Linear Transformation

If X is a normal random variable with mean μ and variance σ2, and

if a ≠0, b are scalars, then the random variable

Y = aX +b is also normal, with mean and variance E[Y] = aμ + b,

var(Y) = a2 σ2.
Summary
N µ 62
P X 1 9 x

P BB 1 9
bf
a 1 d x x
Normal Tables
Example/iClciker roBsedEsoteEfa

• Let X have the N(10, 16) distribution (i.e. Gaussian distribution with
mean 10 and variance 16). Find the numerical values of the following
probabilities: P{X ≥ 15}, P{X ≤ 5}, P{X2 ≥ 400}, and P{X = 2}

µ 1o 22 46 6 4
X 0115
15 P 7
P 1 1

I Q 1.25
Ike 0 1056
a
P X 5 91 1 9 5 0
91
Q
1 25
1 25 0 105
b P 740
164 I t.EE
 17 5 15 a

Q 2 5 0 0062
Q 12 5 Q

c
P X 2 0
Example
• Suppose X is a random variable with mean 10 and variance 3. Find the
numerical value of P{X < 10 − √ 3} (or, nearly equivalently, P{X < 8.27})
for the following two choices of distribution type: (a) Assuming X is a
Gaussian random variable, (b) Assuming X is a uniform random
variable.

a
if X is N 1013 µ 10 6 53

P 210 F P É Xf 1 9
10 51 91 11 1 ans QM
20 1587
b 10 22 3
µ
uniform distribution
3
µ
a 8
14
b a 7 b 13
solve 220 for a

7 13

8 27
8 1 can
P
The central limit theorem and the Gaussian
approximation
• The Gaussian distribution arises frequently in practice, because of
the phenomenon known as the central limit theorem (CLT), and the
associated Gaussian approximation. There are many mathematical
formulations of the CLT which differ in various details, but the main
idea is the following:
if many independent random variables
areadded together and if each of them is small
in magnitude compared to the the
seem then sum
has approximately
a dist
gaussian