Homework/Assignment 8

(Chapter 4: Random Numbers)

Full Name of submitter: Cao Thị Minh Anh Student ID: IELSIU18002
Your Group number is: 2

Group members are:

Student ID Name
1 IELSIU18002 Cao Thị Minh Anh
2 IELSIU18003 Cao Đoàn Trúc Phương Anh
3 IELSIU18068 Trương Ngọc Huyền Linh
4 IELSIU18096 Lương Thanh Bảo Ngọc
5 IELSIU18254 Đinh Thị Hồng Vui
6 IELSIU17041 Lê Phạm Hồng Ân
7 IELSIU18092 Trần Thị Hoàng Ngân

ANSWER
Example 9-5
a. Find the best fit of distributions
Since in the example, the histogram gives the number of vehicles passing through Northwest
corner of an intersection each 5-minute period from 7:00 – 7:05 AM, which is best fit the
physical basis of Poisson Distribution modeling the number of independent events that occur in
a fixed amount of time.
b. Estimate parameters
 Figure 1. Histogram of number of arrivals per period
Total number of observation N = 100
Arrivals per Period Frequency
( X i) ( f i)
0 12
1 10
2 19
3 17
4 10
5 8
6 7
7 5
8 5
9 3
10 3
11 1
100

11 11
2
By calculation, we can find out that ∑ f i X i=364 and ∑ f i X i =2080
i=1 i=1

The sample mean and variance are

364
X́ = =3.64
100
2 2080−100× 3.642
s= =7.63
99

c. Comment about the result

As the result presented above, it’s obvious that with the mean equal to 3.64, the
distribution of the sample is likely to skew to the left and the variance is somehow
acceptable. However, for simulating traffic issues, it is not common that the sample size is
recorded with 100 vehicles, they should have been much larger.
Example 9.7
The observations are now ordered from smallest to largest:
j Value j Value j Value
1 99.55 6 99.98 11 100.26
2 99.56 7 100.02 12 100.27
3 99.62 8 100.06 13 100.33
4 99.65 9 100.17 14 100.41
5 99.79 10 100.23 15 100.47
Parameter estimators:
X́ =99.9 9
s2=(0.2832)2

j
1
2
3
4
 Installation time

99.500 99.600 99.700 99.800 99.900 100.000 100.100 100.200 100.300 100.400 100.500

Figure 2. Q-Q Plot of Installation time

We can see that Installation Times create a straight line, supporting the hypothesis of a normal
distribution.

Figure 3. Histogram of installation time

This q-q plot and histogram of the installation times imposed density function of the normal
distribution.
In conclusion, the door installation times follow a normal distribution.
Goodness-of-fit
Question 1:
α =0.05
x́=1
Let x be the monthly number of job-related injuries.
H0: the monthly number of job-related injuries is Poisson distributed.
H1: the monthly number of job-related injuries is not Poisson distributed.
c n
∑ j=1 f j X j 2 ∑ j=1 f j m2j −n X̄ 2
X̄ = S=
n n−1
c=k=6
s=1
Degree of freedom is k-s-1 = 6-1-1 = 4
c

∑ f j Xj 249
X́ = j=1 = =2.49
100 100
ܱ௜ െ ‫ܧ‬௜ ଶ
X ‫݂݋ ݕ ݑ ݑܿ݊݁ݍ ݁ݎܨݑ‬
݂ ‫ܿܿ݋‬
‫ݑ‬ ‫ ݁ܿ݊݁ݎ‬ǡܱ௜ ݁ܿ‫ݐݔܿ݁ܧ‬
݁݀ ‫ܿ݁݌‬
݁݀ ‫ ݕ ݑ ݑܿ݊݁ݍ ݁ݎ݂ݑ݀݁ݐ‬ǡ ୧
‫ܧ‬௜
0 25 8.290996658 33.6739724
1 47 20.64458168 33.6460232
2 15 25.70250419 4.45651502
3 5 21.33307848 12.5049675
4 3 13.27984135 7.95756029
5 5 6.613360993 0.39358712
Tot al 100 92.6326256
χ 20 =92.633> χ 20.05 , 4=9.488
So, H 0is rejected. The monthly number of job-related injuries is not Poisson distributed.
Question 2:
a. To test the frequency of the data, the proposed technique is Chi-square test since:
- A chi-squared test is used to compare binned data (e.g. a histogram) with another set of
binned data or the predictions of a model binned in the same way
- The large sample size is best fit to Chi-square test rather than K-S test.
 b.

30
24
Frequency of occurrence
25

20

15 12
10 8

5 3 2 1
0
0 1 2 3 4 5
Number of defects per hour

Figure 4. Histogram of the data

n n
∑i =1 X i ∑i=1 X 2i −n X̄ 2
X̄ = =1 . 24 S 2= =1. 288163
n n−1

Frequency of Expected ܱ௜ െ ‫ܧ‬௜ ଶ

X
occurrence frequency ‫ܧ‬௜
0 12 14.4692109 0.42137768
1 24 17.9418215 2.04558531
2 8 11.1239293 0.87729203
3 3 4.59789079 0.55531005
4 2 1.42534615 0.231682
5 1 0.35348584 1.18245344
Total 50 5.31370052

α =0.05
x́=1
Let x be the monthly number of job-related injuries.
H0: the number of defects per hour is Poisson distributed.

H1: the number of defects per hour is not Poisson

Degree of freedom is k-s-1 = 6-1-1 = 4
χ 20 =5.3137< χ 20.05 , 4=9.488
So, H 0is not rejected. The monthly number of job-related injuries is Poisson distributed.
Multiple choice questions
1. Which of the following distribution(s) is for discrete variable (you can choose more than
one)?
a. Exponential
b. Poisson
c. Normal
d. Uniform
e. Weibull
f. Empirical
2. Which method can be used to identify the distribution with data?
a. Histogram
b. Q-Q plot
c. K-S test
d. All of the above
3. Which one is not true about Q-Q plot?
a. To evaluate the fit of the chosen distribution
b. To evaluate if 2 sets of data have the same distribution
c. The linearity of the points in the middle of the plot is less important than the
linearity at the extreme
d. The observed values will never fall exactly on a straight line
4. Which hypothesis testing can be used to test goodness of fit when the number of data is
limited?
a. Chi-square test
b. T-test
c. Pair T test
d. K-S test
5. Which of the below statement is not true regarding the Nonstationary Poisson Process?
a. Arrivals occur one at a time
 b. The distribution of the number of arrivals between t and t+s depends only on the
length of the interval s, not on the starting point t
c. The number of arrivals during nonoverlapping time intervals are independent
random variables
d. b&c
6. Average delay in a queuing system is expected to be 5.3 minutes, and “close enough”
between simulation and real system is 1 minute. A 95% C.I of expected delay based on 10
replications from simulation is [2.5, 5.1]. Which conclusion/ action is recommended?
a. Need to refine the simulation model
b. Accept the simulation model as close enough to be considered valid
c. Additional simulation replications are necessary to shrink the confidence interval
until a conclusion can be reached.
d. No conclusion
7. Which of the following statement(s) is (are) TRUE about histogram?
a. It corresponds to the cdf
b. Its shape is independent with the number of interval
c. It can be plot for either continuous or discrete data
d. All of the statements are FALSE
8. What is (are) LIKELY TRUE concerning the below histogram
a. Need to recheck the data for outliers.
b. Divide data into 2 groups and find a theoretical distribution to fit each group
c. Use an empirical distribution to fit the histogram
d. A complicated distribution like Weibull may fit it.
9. Which Q-Q plot below may correspond with the histogram in question 9

(a) (b)
 (c) (d)
10. Suppose that we want to simulate a quality testing process. Samples from each batch are
drawn out for testing. The probability of finding out a defective product out of n samples most
likely follow which distribution?
a. Binomial
b. Poisson
c. Normal
d. None of the above
11. Which distribution below is NOT appropriate to simulate arrival rate of customers
a. Poisson
b. Exponential
c. Normal
d. Empirical Discrete
12. A data set of 10 intervals is tested for Weibull distribution using Chi-square test. What is the
degree of freedom of the test?
a. 9
b. 8
c. 7
d. 6
Note: dof = k – s – 1 = 10 – 2 – 1 = 7
13. Select the most appropriate statement “Input Analyzer (ARENA) helps to….”
a. input is .p file of simulations runs
b. see which distribution fits the data set best
c. see how change in inputs affect change in output
d. compare 2 different system designs
14. Fill the missing information in the following distribution summary table

Distribution
Name Parameters Application context
 Models the distribution of a process that can be
thought of as the sum of a number of
component processes; for example, the time to
assemble a product which is the sum of the
times required for each assembly operation.
 - μ : mean Notice that the normal distribution admits
- σ: standard negative values, which may be impossible for
Normal deviation process times.
Models the time between independent events,
or a process time which is memoryless (knowing
how much time has passed gives no information
 λ: rate parameter, about how much additional time will pass
which equals to the before the process is complete); for example,
mean number of the times between the arrivals of a large
events in the given number of customers who act independently of
Exponential time interval each other

- k: shape
parameter. This
must be a positive
integer. In the
Gamma distribution,
k can be any real
number, including
fractions.
- μ: the scale
parameter. Must be A patient must go through k steps to complete a
a positive real health screening process. Each step has an
 Erlang number. exponential distribution of service time
  λ: mean number
of events in the The number of random events occurring in a
 Poisson given time interval fixed interval of time
 Triangular Used in situations in which the exact form of the
- a: the minimum distribution is not known but estimates (or
guesses) for the minimum, maximum most likely
value
values are available.
 - b: the peak value
- c: the maximum
value
 - n : total number of
individuals in the
sample
- p :probability of
nonconforming
items produced in The number of nonconforming products in a
 Binomial one trial sampling process
a and b, in which
the minimum value,
a, is the lower
bound, and 0 ≤ a<b.
The maximum
value, b, is the  Models complete uncertainty , since all
Uniform upper bound, and outcomes are equally likely. This distribution is
Distribution a< b<∞ . often overused when there are no data.
 Models the time to failure for components; for
Weibull  α , β are scale and example, the time to failure for a disk
Distribution shape parameter drive.

15. Which hypothesis test formalizes the intuitive idea of comparing the histogram of the data
to the shape of the candidate density or mass function?
a. Chi-square test
b. T-test
c. Pair T test
d. K-S test