Chapter 7

Basic Probability

1
 Introduction
Definitions
▪ A set: any well-defined collection of objects.

▪ Elements: the individual objects that belong to a set.

▪ A subset: a collection made up of some of the objects in a

set .

▪ An event: the subset of results that satisfy a particular

condition or possess a particular property

2
 cont.…

❖The chance of extracting from a whole set of equally

likely and possible outcomes the subset of outcomes

that are favorable is given by
number of favourableoutcomes( success)
Pfavourable =
outcomes number of possibleoutcomes
The range of probability of any outcome is
0  P  1. 0
The sum of Pu and PF is unity
Pu + PF = 1, Pu = Unfavorable outcomes
3
 7.1 Mutually Exclusive Event
(Addition Theorem for Mutually exclusive event)
➢ Events are said to be mutually exclusive if they cannot
possibly happen simultaneously.
➢ Occurrence of one result precludes the occurrence of
any other result.
▪ Mutually exclusive events have no results in common.
▪ Probability that either event 1 (E1) or event 2 (E2) will
occur for mutually exclusive events is:
E1 E2

PE1 or E2 = PE1 + PE2

4
 7.2 Not Mutually Exclusive Events (Addition
Theorem)
➢ When event E1 or event E2 or both can occur
simultaneously, we have not mutually exclusive events.
➢ When several events are involved, at least one of the
events will occur.
➢ Probability that E1 or E2 or both will occur for not
mutually exclusive events is PE OR E = PE + PE − PE and E
1 2 1 2 1 2

E1
E2

5
 Example 7.1

• What is the probability of a die showing a 2 or

a 5?

6
 Cont.…
Ex. 7.2 what is the probability of getting a sum
of 7 or an 11 in rolling two dies?
Solution
 [ (1, 1) (1, 2) (1, 3) (1, 4) (1, 5) (1, 6)
(2, 1) (2, 2) (2, 3) (2, 4) (2, 5) (2, 6)
(3, 1) (3, 2) (3, 3) (3, 4) (3, 5) (3, 6)
(4, 1) (4, 2) (4, 3) (4, 4) (4, 5) (4, 6)
(5, 1) (5, 2) (5, 3) (5, 4) (5, 5) (5, 6)
(6, 1) (6, 2) (6, 3) (6, 4) (6, 5) (6, 6) ]
 So, pairs with sum 7 or 11 are (1, 6) (2, 5) (3, 4) (4, 3) (5,
2) (6, 1) (5,6) (6,5)
7
 Cont.…
Let E1 represents rolling a sum of 7

E2 represents rolling a sum of 11

𝟔
probability of drawing sum of 7 is 𝑷𝑬𝟏 =
𝟑𝟔

𝟐
Probability of drawing sum of 11 is 𝑷𝑬𝟏 =
𝟑𝟔

𝑷𝑬𝟏𝒐𝒓𝑷𝑬𝟐 = 𝑷𝑬𝟏 + 𝑷𝑬𝟐

𝟔 𝟐 𝟖 𝟐
𝑷𝑬𝟏 𝒐𝒓 𝑷𝑬𝟐 = 𝑷𝑬𝟏 + 𝑷𝑬𝟐= + = =
𝟑𝟔 𝟑𝟔 𝟑𝟔 𝟗

8
 Cont.…
Example7.3 A Rivet is to be selected from a box of
1000 rivets that contain defects showen in the table.
a) What is the probability that Number Type of Defect
of Rivets
The rivet will have type A defect
Or type B defect ? 60 A
b) What is the Probability that 35 B
the rivet will have type B defect 15 C
Or type C defect? 8 A&B
C) What is the Probability that
6 A&C
The rivet will have no defect?
3 B&C
2 A&B&C
9
 Solution

P(A) 0.060 =60/1000

P(B) 0.035 =35/1000
P(C) 0.015 =15/1000
P(A n B) 0.008 =8/1000
P(A n C) 0.006 =6/1000
P(B n C) 0.003 =3/1000
P(A n B n C) 0.002 =2/1000

a) Probability of defect A or defect B is

P (A u B) = P(A) + P(B) – P(A n B)
P (A u B) = 0.060 + 0.035 – 0.008 = 0.087
b) P (B u C) = P(B) + P(C) – P(B n C)
10 P (B u C) = 0.035 + 0.015 – 0.003 = 0.047
 Cont…
 C) P(no defect) = 1- P( at least one defect)

P ( at least one defect)=P(A u B u C) =

𝑷 𝑨 + 𝑷 𝑩 + 𝑷 𝑪 − 𝑷 𝑨𝒏𝑩 − 𝑷 𝑨𝒏𝑪 − 𝑷 𝑩𝒏𝑪 + 𝑷(𝑨𝒏𝑩𝒏𝑪)

P(at least one defect) = 0.060+0.035+0.015-0.008-0.006

- 0.003+0.002 = 0.095

P(no defect) = 1-P(at least one defect)=1-0.095= 0.905

11
 7.3 Independent Events
(Multiplication theorem for independent event)
▪ Event E1 is said to be independent of event E2 if the
probability of occurrence of E1 is not affected by the
occurrence or nonoccurrence of E2.
▪ The probability that both events will occur in a trial for
independent events E1 and E2 is PE1 and E2 = PE1  PE2
Example 7.4 Three independent machines are connected in series. The
probability of survival of each machine is 0.98 for 1000 hours. What
is the probability that the failure will not occur in the system for
1000 hrs. of operation?

Solution P𝑬𝟏𝒂𝒏𝒅𝑬𝟐𝒂𝒏𝒅𝑬𝟑 = 𝟎. 𝟗𝟖 ∗ 𝟎. 𝟗𝟖 ∗ 𝟎. 𝟗𝟖 =0.94

12
 7.4 Conditional Probability (Multiplication Theorem)

➢ Event. E2 is said to be conditional on event E1 when the

probability of occurrence of E2 depends on the prior
occurrence of E1.
➢ The probability that E2 will occur provided E1 has
occurred is denoted by ,

PE 2 / E1
which is read as “P of E2 given E1 ”.

13
Cont…

In this conditional probability

- the possible events are those in E1 , and the favorable
events are those in E2 which are also in E1.
number of resultsin E1  E2
PE2 / E1 =
Number of resultsin E1
The probability that both events will occur is

PE1 and E2 = PE1  PE2 / E1

PE2 and E1 = PE2  PE1 / E2
14
  Example 7.5:-A fair die is rolled, Let A be the event that
shows an outcome is an odd number. Also, suppose B the
event that shows the outcome is less than or equal to 3. Then
what is
a. the probability of A, P(A), and
b. what is the probability A given B, P(A|B).
c. What is the probability that both events (P(A) and P(B)) will
occur?

15
 Solution
a) P(A),
 total sample space= 6,
 Total odd number when rolling dice once= 3
 Hence, P(A)= Event A/sample space (S) = |(1, 3, 5)|/S
= 3/6
= 1/2.
b) P(A|B), for calculating conditional probability of A given that B has
happened.
 B has the outcomes {1,2,3} and A has {1, 3, 5}.
 Here (A⋂B)= {1, 3} that are two numbers.
 So, P(A|B)= P(A⋂ B)/ B
= 2 / 3.
c) P(B) and P(A)=P(B) x P (A)/(B)
= 1/2* 2/3
16 = 1/3
 7.5 Probability Distribution
❖A probability distribution is a mathematical
idealization, or model, of the relative frequency
distribution of outcomes of a random experiment.

❖A Probability distribution gives the probabilities of all

possible events that can result from a given trial.

❖Probability distributions are classified as Discrete and

Continuous.

19
 (a) Discrete Probability Distributions

➢ When the events are discrete, i.e. when countable

number of values are assigned to the results of a
random experiment, the probability distribution is
referred to as discrete probability distribution.

➢ Examples of such distribution include: flipping a coin,

rolling a dice where the number of events are finite.

20
 Cont…
 Example 7.6 Determine the probability of obtaining

at least one tail when flipping three coins

simultaneously.

 Solution :- Let P be the probability of flipping a head

and the q be the probability of flipping a tail.

P=0.5,q=0.5

 The probability outcomes and their probability are

shown in the table .

21
 Cont…
Possible Frequency Probability
outcome
HHH p3 0.125
HHT p2q 0.125
HTH p2q 0.125
HTT pq2 0.125
THH p2q 0.125
THT pq2 0.125
TTH pq2 0.125
TTT q3 0.125
Total 𝒑𝟑 + 𝟑𝒑𝟐𝒒 + 𝟑𝒑𝒒𝟐 + 𝒒 𝟑 1.000

the probability of flipping one tail is therefore

𝒑𝟑 + 𝟑𝒑𝟐𝒒 + 𝟑𝒑𝒒𝟐 + 𝒒 𝟑 = 1-p3 =0.875

22
 1.Binomial Probability Distribution
▪ The binomial experiment is any experiment that can be

regarded as a sequence of n trials in which

i. n is defined before the experiment begins;

ii. the result of every trial can be classified into one of

two mutually exclusive categories, success and failure;
iii. the result of any trial is independent of the results of
other trials; and
iv. the probability of success p does not change from trial
to trial.
23
 Cont…
➢ The binomial probability distribution is obtained from

the binomial expansion which is given by

 n  n −1  n  n−k k
( x + y ) = x +   x y +  +   x y +  + y
n n n

1 k 

The n-k coefficient is   =
n n!
 k  k!(n − k )!
 
The binomial distribution is applied to problems where the
probability of occurrence p and probability of non-occurrence q
do not change by drawing from the sample.

24
 Cont…
▪ The binomial expansion may be expressed as

n!
P( r ,n , p ) = p r q n−r , q =1− p
(n − r )!r!
P(r ,n , p ) = probability of r failuresin a sample
of n units when the probability of
the occurenceis p

r = Number of occurrences
p = Probability of the occurrence
n = Sample size
25
 Example 8: A box of candies has many different colors in
it. There is a 15% chance of getting a pink candy. What
is the probability that exactly 4 candies in a box are pink
out of 10?
 Solution
 n = 10,
 p=0.15,
 q=0.85,
 r=4
n!
P( r ,n , p ) = p r q n−r , q =1− p
(n − r )!r!
10!
30
𝑃(4,10,0.15) = 0.154 0.8510−4 =0.04
10−4 !4!
 2. Multinomial Distribution

➢ The multinomial distribution is a generalization of the

binomial distribution in which the occurrence or non-

occurrence of an event can take place in more than two
ways.

➢ For example, in considering an operating unit, it may

malfunction in more than two ways. Such cases are easily

handled by using the multinomial distribution.

31
 3.Poisson Distribution
✓ The Poisson distribution is a probability distribution that

models the number of events occurring within a fixed

interval of time or space, given the average rate of
occurrence.

✓ In applying the Poisson distribution, events are assumed

likely to occur in a continuum of time, length, volume, etc.

✓ - In deriving the Poisson distribution knowledge of the

expected or average number of occurrences of an event is

required.

36
 Example

➢ Let the expected number of occurrences of an event be a.

The expected number of occurrences a can be evaluated from

the probability of an occurrence.

➢ If the probability of an occurrence is p and the sample

size is n, then a = np.
➢ EX.7.9 consider a production line which is normally 10%

defective. If a sample of 50 units is randomly drawn, the expected

number of defects is

Solution a = np = 50 × 0.1 = 5
37
Cont.….
Poisson Formula
Like the binomial distribution, Poisson distribution consists of a
number of probability terms that sum up to 1.0 (The Maclaurin
series expansion for e^x is given by).
2 3
x x
e = 1+ x + + +
x

2! 3!
x −x
ee =1
 x x  2 3
e 1 + x + + +  = 1
−x

 2! 3! 
38
 Cont.…
2 3
−x x −x x −x
−x
e + xe + e + e +  = 1
2! 3!
 Substituting a for x, a being the expected number of
occurrences of an event, we get the following
distribution.
2 3 k
−a a −a a −a
−a a −a
e + ae + e + e +  + e = 1
2! 3! k!

39
 Cont.…
❖Each Poisson term represents probability of an
occurrence.
✓ Poisson term Probability of occurrence of Event
e−a probability of zero occurrence
−a
ae probability of one occurrence
2
a −a
e probability of two occurrences
2!

a k −a
e probability of k occurrences
40
k!
  Example 7.10: :if the expected number of occurrence of
an event a=0.5, what is the probability of zero
occurrences of an event; 1,2,3 occurrences? What is the
probability of up to 2 occurrences?
 Solution
➢ For 0 occurrence 𝒑𝟎 = 𝒆−𝒂 = 𝒆−𝟎.𝟓 = 𝟎. 𝟔𝟎𝟔𝟓
➢ For 1 occurrence 𝒑𝟏 = 𝒂𝒆−𝒂 = 𝟎. 𝟓 𝒆−𝟎.𝟓 = 𝟎. 𝟑𝟎𝟑𝟑
𝒂𝟐 −𝒂 𝟎.𝟓𝟐 −𝟎.𝟓
➢ For 2 occurrences 𝒑𝟐 = 𝒆 = 𝒆 = 𝟎. 𝟎𝟕𝟓𝟖
𝟐! 𝟐!
𝒂𝟑 −𝒂 𝟎.𝟓𝟑 −𝟎.𝟓
➢ For 3 occurrence 𝒑𝟑 = 𝒆 = 𝒆 = 0.0126
𝟑! 𝟑!
▪ The probability up to two occurrence
▪ 𝑷 ≤ 𝟐 = 𝟎. 𝟔𝟎𝟔𝟓 + 𝟎. 𝟑𝟎𝟑𝟑 + 𝟎. 𝟎𝟕𝟓𝟖 = 𝟎. 𝟗𝟖𝟓𝟔

41
Cont.…
 Ex 7.11: What is the probability of no failure for a system with a
constant failure rate of 0.0065 failures/hr. which is expected to
function for an operational period of 100 what is the probability
that more than two failure will occur during the operational
period?
 Solution: the number of expected failure is
 𝒂 = 𝝀𝒕 = 𝟎. 𝟎𝟎𝟔𝟓 ∗ 𝟏𝟎𝟎 = 𝟎. 𝟔𝟓,
 The probability that no failure will occur is given by the first
Poisson term , 𝑷𝒐 = 𝒆−𝒂 = 𝒆−𝟎.𝟔𝟓 = 𝟎. 𝟓𝟐𝟐
 The probability that two or less failure will occur is given by the
three Poisson terms
𝒂𝟐 −𝒂
 𝑷 ≤ 𝟐 = 𝑷𝟎 + 𝑷𝟏 + 𝑷𝟐 = 𝒆−𝒂 + 𝒂𝒆−𝒂 + 𝒆 =0.972
𝟐!
 𝑷 > 𝟐 = 𝟏 − 𝑷 ≤ 𝟐 = 𝟎. 𝟎𝟐𝟖

42
 7.5.2 Continuous Probability Distribution
✓ In a continuous probability distribution, there are an

infinite number of events possible

✓ i.e. a random variable can assume any real value and is

called a continuous random variable.

✓ Many physical properties studied which are continuous in

nature can be described by continuous probability

distributions
✓ For Example , failure of an equipment can be expressed
by using a continuous probability distribution.
43
i. Normal Probability Distribution (Gaussian Distribution)
 The normal curve has the bell shape shown in the figure. The sum
of area under the curve is 1.0.

x
− −  + +

44
 Cont.…
▪ The mid-point µ, the mean of the population, occurs

under the highest point on the curve.

▪ Since the total area under the curve represents all

possible events, portions of the area under the curve

represent the relative frequency of events falling within
this portion.

▪ In determining the significant portions of the area under

the normal curve, the standard deviation is used, which is a

measure of the deviation of the data from the mean.
45
 Cont.…
➢ The standard derivation is determined for two cases:
(i) When grouping is not feasible
(x − x ) x
2

= , x=
n −1 n
(ii) When grouping is feasible
2
 fd
2
fd
 =i −
tot

n  n 
f = frequency of occurrence.
d = deviation from 0 po int .
i = the cell int erval
the 0 point is usually corresponds to the highest frequency
46
Cont.…
Z-Scale
Given σ , it is convenient to be able to read the
horizontal scale in terms of standard deviation
multiples. This is accomplished by converting to the Z
scale by using the formula: x−
Z=


When x =  , the value of Z is zero.

47
 Example 7.12: An equipment which shows a normal distribution
time to failure distribution has a mean of 6720h and a standard
devotion of 1120h . What is the probability that the equipment
may be expected to function without failure: a) for 8400h or more
b) for more than 3900h; c) for more than 3900h and less than
8400h?
 Solution a)using the Z –scale associated with t= 8400h,
𝒕−𝒕𝒎𝒆𝒂𝒏 𝟖𝟒𝟎𝟎−𝟔𝟕𝟐𝟎
 𝒛= = = 𝟏. 𝟓, the area between z=0 and z=1.5
𝝈 𝟏𝟏𝟐𝟎
is 0.4332, the probability that the equipment will function for
t>8400h
 𝑷𝒕 > 𝟖𝟒𝟎𝟎 = 𝟏 − 𝟎. 𝟒𝟑𝟑𝟐 + 𝟎. 𝟓𝟎𝟎𝟎 = 𝟎. 𝟎𝟔𝟔𝟖
𝟑𝟗𝟎𝟎−𝟔𝟕𝟐𝟎
 b) 𝒁 = = −𝟐. 𝟓, 𝒄𝒓𝒐𝒑𝒐𝒏𝒅𝒊𝒏𝒈 𝒗𝒂𝒍𝒖𝒆 𝒊𝒔 𝟎. 𝟒𝟗𝟑𝟖
𝟏𝟏𝟐𝟎
 𝑷𝒕 > 𝟑𝟗𝟎𝟎 = 𝟎. 𝟒𝟗𝟑𝟖 + 𝟎. 𝟓𝟎𝟎𝟎 = 𝟎. 𝟗𝟗𝟑𝟖
 C) 𝑷𝟑𝟗𝟎𝟎 ≤ 𝒕 ≤ 𝟖𝟒𝟎𝟎 = 𝟎. 𝟒𝟑𝟑𝟐 + 𝟎. 𝟒𝟗𝟑𝟖 = 0.9270
48
ii)The negative exponential probability density function (pdf)
▪ A given item is likely to fail in a given time shortly after

installation as in a given time many months latter.

▪ The item is always ‘as good as new’. The cause of failure is

external to the item. In this case, the pdf of time to failure is

given by

f (t ) =  exp(− t ) = e − t

where λ = average failure rate, and

t = time.
49
 Cont.…
The probability of failure in time interval from 0 to t is
obtained from

F (t ) = 0 f (t ) dt = 1 − e − ( t )
t

The reliability, or survival probability for time t is

R(t ) = 1 − F (t ) , or
R(t ) = e − ( t )

50
 Example7.13: A device has a failure rate of 0.5% failures per
1000h. What is the probability (R(t) of the device for an operating
period of 500h? If there are 1000 items in operation , how many
failures are expected in 500h?
𝟓𝒇𝒂𝒊𝒍𝒖𝒓𝒆𝒔
 𝝀 = 𝟎. 𝟓% 𝒇𝒂𝒊𝒍𝒖𝒓𝒆𝒔 𝒑𝒆𝒓 𝟏𝟎𝟎𝟎𝒉 = = 𝟓 ∗ 𝟏𝟎−𝟔 𝒇𝒂𝒊𝒍𝒖𝒓𝒆𝒔/𝒉
𝟏𝟎𝟔 𝒉
 For an operating time of 500h, the probability (R(t))
−𝟔
 𝑹 𝒕 = 𝒆−𝟓∗𝟏𝟎 ∗𝟓𝟎𝟎 = 𝟎. 𝟗𝟗𝟕𝟓 ,
 The number of expected failures for 500h is , 𝑵𝒇 = 𝑵 − 𝑵𝒔
Where N is the total number of device
Nf is the number of failed device
Ns is the number of device that have survived
𝑵𝒔 = 𝑵 ∗ 𝑹 𝒕 = 𝟏𝟎𝟎𝟎 ∗ 𝟎. 𝟗𝟗𝟕𝟓 = 𝟗𝟗𝟕. 𝟓
Therefore, the number of expected failure is
𝑵𝒇 = 𝟏𝟎𝟎𝟎 − 𝟗𝟗𝟕. 𝟓 = 𝟐. 𝟓 = 𝟑
51
 iv) The Weibull probability distribution

▪ Weibull distribution is a useful semi-empirical

expression developed by W. Weibull for use in his
studies of the strengths of steel.
 This has a unique characteristic to adapt to different
situations which is why it is primarily used in
reliability and life data analysis.
 Parameter values affect the distribution which is used
to model the different behaviors for a particular
function.
 Distribution function is generally determined by the
probability density function.

54
 Cont.…
 So, Weibull’s distribution reliability is measured with the help of
parameters
 The scale, shape, and location of the probability density function
are managed by the parameters in the distribution.
 Weibull distribution is considered to be one of the best
methods to calculate life data among several other methods
used to calculate the reliability of data.

55
 Cont.…
▪ In maintenance, the stressing agent is running time t, since new
or since last overhaul.
▪ Weibull failure probability distribution is given as
 
 t −t    t −t 0
−


 
F (t ) = 1 − exp −  0
  = 1− e 

    
where t0 = threshold time-to-failure or guaranteed life
 = characteristic life
β = shape factor.
and

  t − t0  
  t −t0
− 
R(t ) = exp −   =e   

60     
 Example 7.14
A component has a characteristics life (scale parameter)
of 300 hrs. The shape parameter is 2.5. what is its
reliability at 126 hrs if the location parameter is
50hr?
Solution  = 300; t = 50  = 2.5;0

 t −t0 

Given −


R (t ) = e  
2 .5
 126 −50 
− 
R (126) = e  300 

R (126) = 0.968
61
 Cont.….
▪ Example 7.15: Determine the probability of items to fail when
the characteristic life of the items is equal to the interval
between threshold time-to-failure or guaranteed life and the
time at which it can be expected to fail.
t − t0 =  , R(t ) = exp(− 1) = 0.37
F (t ) = 1 − exp(− 1) = 1 − 0.37 = 0.63

In this case 63% of the items will have failed and 37% survived.
In many cases of wear-out, the first failures do not appear until
some significant running time t0 has elapsed. F(t) Arises after t0.

62
 Cont.…
For Weibull probability distribution, the pdf is

dR(t )  (t − t0 )    (t − t0 )
 
 −1 

f (t ) = − = −  
  
dt   
 


And the failure rate is

f (t )  (t − t0 )
 −1

 (t ) = =
R(t ) 

63
 Cont.…
▪ For components that do not have shelf life, i.e. for components
that no deterioration is considered before they go into service,
the threshold time is zero and the Weibull failure probability
distribution is given by
 
 t  t 
− 
 
F (t ) = 1 − exp −    = 1− e
    

and the reliability probability is give by


 t 
 t
− 
R(t ) = exp −   =e  

    
64
 Example 7-16: Assume that the life of a packaged magnetic disk
exposed to corrosive gases has a Weibull distribution with a
characteristics life of 300 hrs and shape factor 0.5.
Calculate the probability that
a) a disk fails before 500 hours.
b) a disk lasts at least 600 hours,
Solution:
a) P (t<= 500)= F(500) = 1-exp^(-(500/300)^0.5)
=0.725

b) P (t>= 600) = 1- P (t<600)= 1-F(600)

= 1- {1 – exp ^(-(600/300)0.5}
65 =0.2431
 Cont.…

For  =1, and t0=0, the failure rate depends on β and is

given by
 (t ) =  t
 −1

The relationship between the failure rate and β is shown

in the figure below.
Failure
rate, (t) 


=

Time in service, t

66
 Thank you

67