Reference no: EM132069448

THE FLINDERS UNIVERSITY OF SOUTH AUSTRALIA

College of Science & Engineering

STAT8721 Data Science Major Assignment 2. 30%

You are required to present your research in the final week of semester. You will be assigned a time and all group members must attend and present.

Requirements

• Demonstrate an understanding of good data science practice.
• Conduct data analysis to identify and test potential hypotheses.
• Present the results of data analysis in written, graphical and verbal forms.
• Develop and demonstrate your presentation skills.

The aim of this assignment is to encourage you to conduct evidence based research, critique such evidence and present scientific results in a range of standardised forms.

You will work in the same group as you did for assignment 1.

Assignment scenario

I am a supermarket manager wanting to improve business performance and profits. I have several data sets that have been used for previous data analysis. I have provided all available data sets for you to use but you will need to combine them. I am interested to know more about one area of business operation. Based upon your group number, you are asked to focus on and present a report that will provide evidence upon which my board members and I can base our business decisions in response to one of the following questions.

Student  3. How I can sell more products?

Task 1 - Create a data set and develop a hypothesis (Group).

1. Download and combine all supermarket data sources from FLO. Submit as a single data file.

2. Report on issues in linking and combining data. What problems were presented and how did you overcome them? How much data did you exclude and how much is available to use? What is the potential impact of the data loss?

3. Run data mining association rule analysis on your combined data set to find patterns that can form hypotheses. Discuss the patterns found and select one pattern to analyse further. Rewrite the

pattern as a hypothesis. Explain and justify your selection for hypothesis. Note that a null hypothesis is not valid in this assignment.

Task 2 - Test your hypothesis and report findings (group).

1. Develop and appropriately present summary statistics, and any other methods you need, to suggest if the hypothesis should be accepted or rejected. Justify all methods and discuss all results in the context of the hypothesis. Make a clear statement regarding support for the hypothesis.

Task 3 - Present findings to the board (group).

1. Write a 3 page executive summary for the 'client' on what you did (your methodology) and your findings. This should present the outcomes of the work completed in task 2. Provide a clear statement regarding how the business should/could use or apply your findings. Align your report to the business focus area and your hypothesis.

2. To support your report you are required to present your findings as a company board presentation in the final week of semester. Each presentation should be no more than 7 minutes and follow a power point presentation of no more than 5 slides.

3. Each group will present to the class members and guests and convince them of the value of the analysis and findings.

Task 4 - Individual analysis and evaluation

1. Explain how, if you had more time, you would extend the research you have conducted. What were the key issues encountered by your group and how would you overcome/manage these issues in future work? Provide a short discussion of the work you contributed to the group report. What do you feel you as an individual did well and where do you feel you could improve?

Submission

As a group you must submit the following files to FLO by the deadline.

1. A single file containing the combined data set used for analysis.

2. A single, PDF file listing all group members and reporting on the activities and outcomes for task 1 and your executive summary resulting from task 3 separated by a business appropriate title page. Maximum of 3 pages for each part excluding title pages. (Only listed group members will receive a grade.)

4. A single power point file containing the slides for your board presentation.

NOTE: Each group will present in the final week of semester and use this to gain feedback to revise their final submissions.

As an individual you must each submit a single PDF file containing your individual analysis of no more than 2 pages.