Professional Documents
Culture Documents
CHAPTER 2
2.1. Probability
• Definition
• The Venn Diagram
• Operational Rules
• Mutually Exclusive Events
• Conditional Probability and Bayes’ Theorem
• Statistical Independence
Definition
Probability is a measure that reveals the likelihood of occurrence of an event relative to
the set of all possible events in a sample space.
ℎ
, =
ℎ ℎ
Example 2.1: For a simple example of tossing a coin, find the probability of having at
least one head in a sequential two tosses. (H: Head, T:Tail)
Solution:
HH
H
HT 3
Outcome tree = = 0.75
TH 4
T
TT
For the below diagram, and are the defined events of the sample space,
i.e. subsets of the sample space. They are shown by a closed region within the
rectangle.
Sample Space
A certain event, denoted , is the event containing all the sample points in a sample
space. Therefore, it is the sample space itself.
Ex: Assume that the sample space is the diesel consumption amounts of a specific
mining truck.
= Possibility of the mining truck consuming a single drop of diesel or more when
driving.
MinE424 - Project Management in Mining 6
2.1. Probability
XX YX ZX XX YX ZX
XY YY ZY XY YY ZY
XZ YZ ZZ XZ YZ ZZ
⊃ , ,
⊃ , ,
The probability that exactly one job will be completed definitely in one year:
⊃ ,
⊃ ,
Operational Rules
∪ union means «OR»
Important!
∩ intersection means «AND»
⊃ belongs to, or is contained in
⊂ contains
complement of
For ∀ cases, 0 ≤ ≤1
Complementary Event → ( ) = 1 −
Union Rule → ∪ = + − ∩
Intersection Rule → ∩ = | ∗ = | ∗
Operational Rules
Associative Rule → ∪ ∪ = ∪( ∪ )
De Morgan’s Rule → ∪ = ∩
∩ = ∪
( ∪ )∩ = ( ∪ ) ∪ = ( ∩ )∪
.
.
.
When playing with operators and converting them into many different forms,
we need to phrase their meanings with words for the each form. In any
confusion, you may draw a small Venn diagram to validate.
MinE424 - Project Management in Mining 10
2.1. Probability
Operational Rules
Example 2.3: The water supply for a mine comes from two different sources with two
separate pipelines. These pipelines are jointed in a station 30 kms away from the mine
license area and the water is conveyed in a single pipeline thereafter. Explain the
probability of that no water shortage will take place .
Solution: Source 1
3 Mine
Source 2
If the occurrence of one event make the occurrence of other event impossible, they are
called mutually exclusive events. Therefore, mutually exclusive events cannot take
place at the same time.
Ex:
• Making right turn and left turn at a street intersection
• Flood and drought of a river at a given instant time
• Failure and survival of a building to a strong earthquake.
If there are multiple events that are mutually exclusive, then their intersection will be an
impossible event.
∩ ∩ ⋯∩ = ( ∩ ∩ ⋯∩ )=0
MinE424 - Project Management in Mining 12
2.1. Probability
If , , ,… ℎ ℎ , i.e. ( ⋂ ) =
0 ≠ ,
• ∪ ∪ ∪ ⋯∪ = + + + ⋯+
• | =0 | =0
∩ = | ∗ = | ∗
∩
Probability of assuming has occurred | =
∩
Probability of assuming has occurred | =
MinE424 - Project Management in Mining 14
2.1. Probability
∩ = | ∗ = | ∗
| ∗
| =
a) What is the probability of successfully controlling air pollution in the next 3 years?
b) If, in the next 3 years, the pollution level is not sufficiently controlled, what is the
probability that it is entirely caused by the failure to control mineral processing plant?
c) If the pollution is not controlled, what is the probability of mineral processing plant is
not controlled?
Solution: Let’s assume that controlling mining trucks and mineral processing plants are
denoted as A and B, respectively.
( | ) . ∗ .
b) = = = 0.32
( ) .
c) = ∪ ̅ )
= + ̅
∗ ( ) ̅ ∗ ( ̅)
= +
( ) ( )
. ∗( . ) . ( . )
= + = 0.84
. .
= = 0.70
̅ = = 0.30
A laboratory want to perform an additional test for the quality assurance. The test is
not perfectly reliability and good-quality aggregate will pass the test with a probability
of %80 whereas it is %10 for the poor-quality aggregate.
Statistical Independence
If occurrence of one event doesn’t affect the occurrence probability of other event, the
two events are statistically independent.
These events should be confused with the mutually exclusive events since statistically
independent events can take place at the same time but they don’t have any mutual
effects on their occurrences. On the other hand, mutually exclusive events cannot
happen at the same time.
Statistics Probability
(Algebra of Uncertainties)
Uncertainty
Aleotory Epistemic
Random Uncertainity Systematic Error
} Cannot be reduced } Due to lack of sufficient knowledge
Inherent Uncertainity Sampling Error
Measures of Distributions
∫ = 1.0
Probability for the given intervals, ∫ : The area under gives the
probability of event to take place between the values of and .
MinE424 - Project Management in Mining 21
2.2. Probability Distributions
Measures of Distributions
Measures of Distributions
Median: Mean (Expected Value):
= 0.5 =
3 3 10
= = | = 7.5
1000 1000 0
Normal Distribution
Many real life problems and statistical approaches have normal distribution.
Mean=Mode=Median
( )
=
( )
= ∫
=∫ =∫ ( ) = , ℎ = ∑
= ∑ ( − )
Important Points!
Std(x), = Coef. of Variance, = ⁄
Normal Distribution
Example 2.6: The below table gives the compressive strength results for 20 different
limestone core specimens taken from the same mine site. Assuming that the obtained
data fits into normal distribution, find the mean, median and standard deviation
statistics.
30 33 34 31 33
29 30 35 34 30
36 33 35 32 31
29 36 35 32 32
Solution:
1 1
= = = 30 + 33 + ⋯ = 32.5
20 20
1 1
= ( − 32.5) = (30 − 32.5) +(33 − 32.5) + ⋯ = 4.85 ; = 2.20
20 20
MinE424 - Project Management in Mining 25
2.2. Probability Distributions
Normal Distribution
From Example 2.6:
68.3%, 95.5% and 99.7% in the figure are well-known values which are constant for
any dataset in normal distribution. They also refer to confidence intervals. For the
given example, the compression strength values are between 28.1 and 36.9 for the
confidence interval of 95.5%. In other words, the probability of that the strength values
are between 28.1 and 36.9 is 0.955.
MinE424 - Project Management in Mining 26
2.2. Probability Distributions
Normal Distribution
From Example 2.6:
a b
It is very challenging to integrate
this equation if the probability is
1 ( )
checked between the values ( , )
≤ ≤ = =
2 other than the well-known values. In
that case, we may use standard
normal distribution for easiness.
−
= = (30−32.5)/2.2
Actual Numbers Normalized Numbers
= 32.5 ; = 2.20
Then, the probabilistic values for normalized numbers can be read from the general-
purpose z-score table.
1 ( )
= .
2
=
=
1
= . .
2
μx= 60 mm (mean)
σx= 15 mm (standard deviation)
1
P ≤ 80 = Difficult to solve the equation
2 15
P z ≤ 1.33 = 90.82%
. % > 85%
μx= 60 mm (mean)
σx= 15 mm (standard deviation)
1
P 50 ≤ ≤ 80 =
2 15
P ≤ ≤ = P(−0.67 ≤ z ≤ 1.33)
Lognormal Distribution
. .
Variable ln
Expected Value
Standard Deviation
( ( ) )
1 = (ln )
| , =
2
= ln(1 + ) = ln(1 + )
≤ = ≤ = (ln )
1
= ln − = ln = ln
2 1+
Lognormal Distribution
Example 2.9: Estimate the probability in Example 2.8 by converting the data into
lognormal distribution.
Solution:
= 60
= 15
1
= ln 60 − 0.25 = 4.06
2
60
= ln = 4.06
1 + 0.25
ln 50 − 4.06 ln − ln 80 − 4.06
P ln 50 ≤ ln x ≤ ln 80 = P ≤ ≤
0.25 0.25
Binomial Distribution
If the probability of occurrence of an event in each trial (probability of nonoccurrence
is 1 − ), then probability of exactly occurrences among trials:
P = = (1 − )
Example 2.10: Alacer Gold Mining Co. uses four graders for the haul road maintenance
of Çöpler Gold Mine. The chief engineer asks the project department to calculate the
probability of at least three graders to be available after 900 hrs. Each grader has the
same operational life with a log-normal distribution with a mean life of 1500 hrs and a
COV, i.e. coefficient of variance , of 30%.
Solution:
= 0.30 ⇢ = = 0.30
1
= ln 1500 − 0.30 = 7.27
2
MinE424 - Project Management in Mining 38
2.2. Probability Distributions
Binomial Distribution
Solution(cont’d):
ln 900 − 7.27 ln −
P 900 ≤ =P ≤ = P −1.56 ≤ = 0.9406
0.30
Probability that there will be at least three grader to be operated after 900 hrs:
4 4
P ≥3 =P = 3 + P( = 4) =
0.9406 (0.0594) + 0.9406 (0.0594)
3 4
P ≥ 3 = 0.1977 + 0.7827 = 0.9804
= +
: Dependent variable
∑( − ̅ )( − ) Correlation Coefficient:
=
∑( − ̅ ) ∑( − ̅ )( − )
=
where ∑( − ̅ ) ∑( − )
∑ ∑
̅= , =
This coefficient defines the
correlation between
= − ̅ dependent and independent
variables mathematically.
34
32
Production Cost ($/ton ore)
30
∑
28 ̅= = 433.5
26
24
∑
= = 28.1
22
20
350 370 390 410 430 450 470 490 510
Evelation Difference (meters)
( − ) ( − ) ( − ) ( − ) ( − )( − )
∑( − ̅ )( − ) 735.9
= = = 0.075 = −4.4 + 0.075
∑( − ̅ ) 9747.8
= 600
= − ̅ = −4.4
= 40.6 $/ton ore
∑( − ̅ )( − ) 735.9
= = = 0.98 ( )
∑( − ̅ ) ∑( − ) 57.7 9747.8
Note: Correlation coefficient varies between -1 and +1. If its value converges to +1, this
means that there is a strong positive correlation between the variables such that any
increase or decrease in the dependent variable can be expressed with almost same
increase or decrease rate of change in the independent variable. If the value converges
to -1, there exists strong negative correlation. In that case, positive change in the
dependent variable can be defined with the almost same rate of negative change in
independent variable or vice versa. A correlation coefficient with a zero value points
that there is not any correlation (dependency) between the variables.
= + +
: Dependent variable
∑ ∑ ∑
= = =
( − ) + ( − )( − )= ( − )( − )
( − )( − )+ ( − ) = ( − )( − )
= − −
Solution:
= 399 = 385 = 1,015 = 22,521
= 57 = 55 = 145 = 56,518
( − ) + ( − )( − )= ( − )( − )
( − )( − )+ ( − ) = ( − )( − )
∑
= ( − ) = − = 1,536
∑ ∑
= ( − )( − )= − = 576
∑
= ( − ) = − = 390
∑ ∑
= ( − )( − )= − = 2,257
∑ ∑
= ( − )( − )= − = 893
Equation-1
= − − = 145 − 57 − 55
Equation-2
Equation-3
( − )( − )+ ( − ) = ( − )( − ) 576 − 390 = 893
=52.23
=1.37
=0.27
Nonlinear Regression
Examples: = , = , etc.