Data Analysis and Visualization Coursework 2

Page | 1
M31995 & M32274-DL Data Analysis and Visualization
Coursework 2
This is an individual assignment and carries 50% of the module mark. It covers the
following learning outcomes for the module:
• Understand and critically apply common algorithms given sets of data
and requirements
• Understand different stages of data science
• Understand fundamental concepts in data analytics and machine
learning
• Understand the process of data cleansing
• Be able to visualise data effectively

There are two sections in this coursework. Under Section A, you are expected to
answer all the questions in two independent tasks, which account for 80% of this
coursework (therefore 40% of your final module mark). Under Section B, you are
expected to write a short report of no more than 300 words, which accounts for 20%
of this coursework (therefore 10% of your final module mark)
Section A and B of this coursework assignment can be found on Page 5 and 7,
respectively.
Submission Hand in Arrangements: Students are required to submit their
coursework assignment to Turnitin via the link on the module Moodle. Students are
encouraged to submit drafts through the Turnitin student checkpoints prior to the
submission deadline.
There are mark penalties for late hand in – see “Late Submission of Coursework”
above. (Corruption of computer files is not an adequate excuse for late hand in, as
work should be adequately backed up.)
PLEASE NOTE that the deadline for submission is 11.55pm on the submission
date – this means that your assessment must be fully uploaded by 11.55pm.
Assessments uploaded at 11.55pm or later will be marked as late. Students
should submit in advance of the final deadline wherever possible. Computer
uploading delays and other IT difficulties are not accepted as extenuating
circumstances.
If there is a problem with Moodle at the time of submission students should
email their coursework to:
Student Number
Module Name / Module number
Course name that you are studying
2 | Page
Screenshot of the error
Word Limit: For Section A: 1500 words excluding footnotes and bibliography but
including all in-text citations (if there is any). For Section B: 300 words excluding
footnotes and bibliography but including all in-text citations (if there is any). An abstract
is NOT required. The word count should be stated at the top of the assignment.
Failure to state a word count will result in a penalty of 5% of the original mark awarded.
A falsely stated word-count is an assessment office which may result in a penalty,
including the reduction of the mark to 0%. Note, that footnotes should be used to
reference sources only. Examiners are free to disregard footnotes that contain
inappropriate information or information that should belong in the main text.
Coursework that is over the stated word limit will result in a penalty of 10% of the
original mark awarded. For the avoidance of doubt, the penalty will be applied to
any work that exceeds the stated word limit of (1500, 300) words excluding cover
page and bibliography. Students are NOT permitted to exceed the word limit by
10% or any other amount, however small. This means that any assignment of
(1501, 301) words or more will be penalised.
Referencing requirements: Students must reference all sources using the APA 7th
Edition. Guidance on this method of referencing can be found at
www.referencing.port.ac.uk. Reference should be made to the primary source, except
when the primary source can no longer be obtained. Poor citation of sources will result
in a loss of marks.
Referencing is required to give intellectual credit to your source, help your reader
recover your source easily and to avoid being accused of plagiarism.
Students are reminded that the University will not tolerate academic dishonesty in any
form. This is cheating.
Plagiarism: Students are reminded of the need to avoid plagiarism. The University
Regulations describe plagiarism as:
the incorporation by a student in work for assessment of material which is not
their own, in the sense that all or a substantial part of the work has been copied
without any adequate attempt at attribution, or has been incorporated as if it
were the student’s own when in fact it is wholly or substantially the work of
another person or persons.
Any student suspected of plagiarising will be referred for an Academic Misconduct
Hearing.
Students should ensure that all sources are fully cited in footnotes and in the
bibliography and that indentation or quotation marks (as appropriate) are used when
quoting. Failure to include a bibliography will result in a 5% penalty, unless the
lecturer/tutor has advised you that a bibliography is not required.
Formatting: The work should be word processed. Font size should be between 11
and 14 and ‘easy to read’ e.g., Calibri, Arial, Times New Roman. Line spacing should
be between 1.5 and 2 with (approx.) 2.5 – 4 cm margins all round. The Header must
include the student number and the Footer must include a page number.
3 | Page
Electronic Copy of Work: Students should retain an electronic copy of their
coursework, so that it may be checked by a member of staff should a member of staff
feel the need to do so. Tutors are entitled to request an electronic copy of coursework
if they have any doubt about the accuracy of the stated word count and/or any
suspicion of plagiarism. Failure to send an electronic copy of the coursework to a
member of staff who has asked for a copy may result in a penalty.
If any student has a query about any of the above matters and wishes to obtain
clarification or further information, please contact your Module Co-ordinator or
personal tutor.
Feedback: Marking will be done in accordance with the marking criteria grid below
and the University of Portsmouth grading criteria for PG level 7 (See next page). Marks
and feedback will be available within 20 working days after assessment deadline.
4 | Page
Marking Criteria (This is a General criteria applicable to essays, reports and aspects
of projects and dissertations)

Level 7
80+ As below plus:
Potentially publishable work
Excellent work – able to express an original reasoned argument in a lucid manner by reviewing & critiquing a wide range
of material. Original, critical thinking based on outstanding insight, knowledge & understanding of material. Material
contributes to current understanding & is of potentially publishable quality in terms of presentation and content
 Wide reaching research showing breadth & depth of sources
70-79
60-69 As below plus:
 Clear, balanced coherent critical & rigorous analysis of the subject matter. Detailed understanding of knowledge &
theory expressed with clarity
 Extensive use of relevant & current literature to view topic in perspective, analyse context & develop new
explanations and theories
50-59 As below plus:
 Detailed review and grasp of pertinent issues & a critical contextual overview of the literature. Thorough knowledge of
theory and methods & uses this to underpin arguments and conclusions
 Confidence in understanding and using literature
40-49  Demonstrates grasp of key concepts & an ability to develop & support an argument in a predominately descriptive way
with valid conclusions drawn from the research
 Familiarity with key literature which is cited and presented according to convention
 Logical & clear structure, well organised with good use of language and supporting material
30-39
0-29
FAIL Some knowledge of relevant concepts & literature but significant gaps in understanding and/or knowledge. Little
attempt at evaluation, conclusions vague, ambiguous & not based on researched material. Limited or inappropriate
research. Deficits in length, structure, presentation &/or prose
5 | Page
SECTION A (80%):
There are two independent tasks in Sections A. You are expected to complete all
questions. You are not expected to write more than 1500 words in total for Section A,
and it can be less. Screenshots from Excel/STATA do not count.
TASK 1:
The data for Task 1 is presented on the sheet “Real Estate” of the excel file
“Coursework assignment two data.xlsx”
You are a business analyst. Your customer, a real estate company, has given you a
sample of observations of real estate prices from their database. Your customer wants
to know how to determine the price of a house once he or she knows the transaction
date, house age, distance to MRT stations, and the number of convenience stores
nearby.
Question 1. Build a linear regression model to predict house value. Interpret the
regression results to your customer. (14 marks)
Question 2. One of your colleagues suggests that a linear model may not be good
enough and he suggests using a nonlinear model. What would you say? If you do not
agree, please explain your logics with relevant references. If you agree, please
suggest a decent non-linear model and show evidence. (10 marks)
Question 3. Assuming that a linear model will be used for this case. One of your
colleagues suggests using a regularised model. Why would she propose using a
regularised model? Could you apply such a model and briefly explain how it deals with
the potential problem(s)? (9 marks)
Question 4. Assuming that a linear model will be used for this case. Another of your
colleagues suggests conducting a Principal Component Analysis (PCA). Do you
agree? Critically evaluate this suggestion. (7 marks)
TASK 2:
The data for Task 2 is presented on the sheet “Iris” of the excel file “Coursework
assignment two data.xlsx”
You are trying to determine the type of Iris (versicolor or virginica). You have obtained
a sample of Iris. This dataset includes the sepal length, sepal width, petal length, petal
width, and the labels.
Question 1. Use the K-nearest neighbour (k-NN) algorithm with the Euclidean
distance measure to determine the type of the following observation (6.5, 3.0, 5.0,
1.4). Use relevant theories of the algorithm to justify your decision. (8 marks)
6 | Page
Question 2. When we classify an unseen instance (flower) by using the k-NN
algorithm in the above dataset (i.e., binary classification), what might be the problem
if we set the value of k to an even number, and what could be a possible solution? (4
marks)
Question 3. Apply 3-NN, 5-NN, 7-NN, and 9-NN models, calculate the confusion
matrix, and evaluate the models. Which model do you think is the best? Explain your
choice. (14 marks)
Question 4. Are you aware of other classification models/algorithms that can be used
to decide the type of Iris? Propose two different types of models / algorithms and
explain briefly how you plan to use them. (14 marks)
80 marks in total for Section A
7 | Page
SECTION B (20%): Data Visualisation
You have studied several aspects of effective charts including the following:
− Can answer most of your questions.
− Help compare and rank.
− Show the correlation between different measures or variables.
− Emphasise the most important data.
− They are not overloading views.
− Show a limited number of colours and shapes in a single view.
Besides, there are several questions to ask before producing effective charts
such as:
− What are you trying to say? What is the story that you want to tell?
− Who is your audience, their experience to understand the graphs?
− What questions might your audience have?
− What answers are they searching for?
− What is an appropriate chart type to meet your audience needs?
− Can your audience understand the visualisation in 30 seconds or less, without
additional information?
The question listed below is NOT relevant to the above SECTION A
Requirements:
Write a report of no more than 300 words to critically evaluate the effectiveness
of a chosen chart (any chart of your choice) in the light of the aspects of effective
charts and the questions discussed above.
The chosen chart should be informative and professional enough. You
can search for those charts produced and published by big companies such as
Amazon, IBM, etc… and choose one to critically evaluate its effectiveness.

20 marks in total for Section B