IERG3050_2021Fall_Assignment5
IERG3050 Simulation and Statistical Analysis (Fall 2021)
Assignment 5
Instructions:
1. Do your own work. You are welcome to discuss the problems with your fellow
classmates. Sharing ideas is great, and do write your own programs and explanations.
2. All work should be submitted onto the blackboard before the due date.
3. You are advised to submit a zip or rar containing all necessary files. In the submission,
you should have the following items
a. Three python programs, assignment5_taskX.py, the program written for the tasks.
(X = 1, 2, 3, 4, 5 or 6)
b. Some .csv files and a .pdf file, explaining your own test case in the extra features.
4. Do type/write your work neatly. If we cannot read your work, we cannot grade your work.
5. If the submitted programs cannot be run or are irrelevant to the problem (e.g. empty
python programs), you will receive no scores for that part.
6. Assignment 5 is equivalent to project phase II, and will take 10% of the course total.
7. Do spend at least three days for the project-like assignment 5.
8. Due date: 10th Dec, 2021 (Friday) 23:59
In assignment 2 (project phase I), we have written a simulation for the CUHK Art Fair. In this
part, we will carry out different analyses based on the results of assignment 2. In other
words, if you did not complete assignment 2 before, you will need to go back and work on
assignment 2 before you can start assignment 5.
In assignment 5, you are going to work on 3 tasks (upon your choices) out of 6 tasks. For
each task, you need to:
1. Describe your experiment design in the document
2. Perform calculations, if any
3. Show the results in the document
4. Give conclusions in the document
5. Submit the program as assignment5_taskX.py, X = 1, 2, 3, 4, 5 or 6
In the submitted document (.pdf file), you need to include the following items:
1. Your student name and student ID
2. Your selected tasks (3 tasks out of the 6 options)
3. Some screenshots and captions explaining how you complete the tasks
4. Some calculations, if any
1
Task 1: Analysis on the total waiting time using 2-stage sampling
This is an extension to the extra features, reordering the preference list based on distance
and reordering the preference list based on queue length, in assignment 2.
[System 1] Reordering the preference list based on distance: As the University Mall is
outdoor and the weather is very hot, students visit the closest booth according to their
preference list.
[System 2] Reordering the preference list based on queue length: If there are too many
people crowding at some booths, it will take a long time to wait. Students visit the booths
with the shortest queue length first.
Which system, system 1 or system 2, has a better performance, a shorter total waiting time?
Explain your answer with 2-stage sampling at 5% significance level.
Task 2: Grouping behavior
This is an extension to the extra feature, promoting to groups of students, in assignment 2.
Promoting to groups of students: The society can promote to at most 6 students at a time. It
is noted that late comers wait in the queue and join the next promotion.
You need to perform the simulation with an extra feature:
After a group of students listening to the promotion of a booth. Some of them may share a
common interesting society. There is a chance, a parameter q determined by you, that these
students group together and visit the common interesting society together. Justify your
simulation results with your chosen parameter q.
Task 3: Workflow representation and traffic generation
What are open class and closed class? Explain these terms using the Air Fair simulation as
an example.
A theorem states that open representations are pessimal. Verify this theorem experimentally.
Task 4: Sensitivity analysis on the total tour time
In the simulation, the output variables are the total waiting time, the total traveling time and
the total tour time. What are the input variables of the simulations? Design the experiments
for determining all variables that the total tour time is sensitive to.
2
Task 5: Multi-linear regression model on the total traveling time
Construct a multi-linear regression model for the total traveling time. It is assumed that the
total traveling time depends on the number of booths in the Art Fair, expected promotion
time of a booth and the expected number of booths visited by a student only.
Task 6: Graphical visualization
Write the graphical visualization in 2D for the simulation. You can use simple shapes, such
as circle, square, rectangle, rhombus and triangle, to represent different objects, such as
booth, student, queue, entrance and exit, for the simulation.
You don’t need to carry out any analyses for this task, but your simulation should be general
enough such that we can see the simulation by supplying different configuration files.
Grading criteria:
Item Description Score
Submission of
document
Some items of the document are missing, or the submitted
document is not a .PDF file.
0
The submitted document contains all necessary information,
and the document is a .PDF file.
1
Each task The task is unattempted, or the submitted work is far from
completion.
0
The logic flow can be understood, however, most parts are not
done.
1
Some errors are spotted, or some parts are unclear. 2
The experiment design is explained clearly. Some test cases
are provided. The results of the analysis are clearly shown and
discussed.
3
3