Assignment: Data Science Project Proposal
For this assessment task, you are required to write a report and prepare a 5 minute presentation regarding a chosen business and data proposal. A draft of the first two sections of your report is due week 8, the final report is submitted in week 11, and the slides (consisting of a maximum of 5 slides) week 12 before lecture. The 5 minute presentation is given in week 12 during your assigned tutorial. You will also be required to review and provide feedback on presentations of other students (peer-review) during the final tutorial in week 12.
Your report should have at least five sections:
· Project Description: provide a description about the data science project that you study/propose, what the project is about, and what data science roles are involved in these project and what are their responsibilities.
· Business Model: provide analysis about the business/application areas the project sits in, what are the challenges of the project, what kind of values the project can create for the specific business area, etc.
· Characterising the Data and Data Processing: characterise the data in the project (i.e., the V's), provide analysis on the required technologies for data processing according to the specific data characteristics.
· Resources: locate and assess potential resources, software and tools for the project. Especially, what are your data sources?
· Data Analysis: specify/propose the statistic methods used in the project, provide analysis on why you chose those methods and discuss about the high level output.
Propose completely new data science work for government sectors in Melbourne.
Note, if we really like your project ideas, we may ask your permission to bo
ow some parts for an Industrial Experience project the following year.
Before you begin
Before starting your project, you should:
· Read the business and data case study example topics and propose your own topic.
· Have a look at examples of good past student reports .
What you need to do and due dates
Choose a topic and then:
· Write the initial two sections (Project Description and Business Model) of the report and submit them first. It should be upto 1000 words.
· Prepare the rest of your report (maximum 3000 words total).
· Submit an exact copy of your report to Turnitin as well as through the assignment submission page. You can submit earlier copies to check if they trigger Turnitin's database. You can submit updates too. But you have to do a final submission before the deadline.
How you will be assessed
Clear description of the goals of the project, appropriateness of topic, clear description of the business benefits, novelty/creativity, overall clarity of the initial report.
Assessment will be assessed on your ability to:
· analyse the role of data in the business model
· discuss different parts of data science project from the perspective of the data science process and from the perspective of the roles such as statistician, archivist, analyst and systems architect.
· analyse the size and scope of data storage and data processing, and present the basic technologies in use.
· locate and assess resources, software and tools for a data science project.
· discuss the kinds of data analysis and statistical methods suitable for the data science project
· think critically and creatively, provide justification and analysis;
Word limit
The maximum word limit for the final report is 3000 words. References at the end of the report (consisting of a list of URLs and/or cited reports) are not included in the word count.
Further advice on the assignment:
Here is some further advice from the lecturer and tutors regarding the assignment:
1. Make sure to carefully read the assignment specification above.
2. The project should be data-centred -- ideally combining multiple sources of data to develop a predictive model that can solve a real-world problem.
3. The project should contain a clear statement of the problem being tackled. What is the objective/purpose of the project? Have a look at the structure of the case studies in the NIST document -- each one starts with a clear definition of the problem.
4. Make sure that the benefit of the project is clear. What is it? Will the project have a financial benefit, or result in a social good?
5. The report needs to be "telling a story", and to be convincing somebody to "invest in your project" so that it can be built.
6. Try not to make the project too
oad. -- It should be an achievable data science project.
7. Include graphics to support claims. For example, depending on your project it may (or may not) make sense to include:
· an influence diagram showing what data is available and how it relates to the decisions and objectives of the project,
· graphs showing some exploratory data analysis (if applicable).
8. Make sure you understand the difference between business models and data models. -- They are not the same thing!
9. Read up as much as you can on the particular topic you've chosen in order to be able to describe the data (and software) requirements of the project.
10. Make it clear where the data would come from for the project:
· Is the data proprietary? How would it be collected?
· If the data is public, you could even do some exploratory data analysis on it.
11. What processing would be needed? How would the data need to be processed before it can be used? What software might be needed? Can the processing be distributed?
12. Finally, make sure you've seen the set of possible section headings suggested above and had a look at the examples of reports from past years. (Note though that the assignment specifications were different last year.)
Microsoft Word - Assessment task 3.docx
WATER RESOURCE ANALYSIS IN
MELBOURNE
Semester 1, 2016
FIT5145 – Introduction to Data Science
Lecturer: Wray Buntine
Tutor: Asef Nazani
Assessment Task 3 – Business and Data Case Study Report
Student: XXX
ID: XXX
ABSTRACT
Since 2012 Melbourne’s water storage levels can be topped up by water orders from the desalination plant. By
using predictive water storage modelling a sensitive decision about placing an order and its height can be
eached. For this data like streamflows and storage levels, meteorological predictions and water consumption
forecasts has to be collected and taken into account. All of this is part of an intricate data science process.
TABLE OF CONTENT
Abstract .................................................................................................................................................. 2
Table of Content ..................................................................................................................................... 2
1. Project Description ...................................................................................................................... 3
1.1. Background .................................................................................................................................. 3
1.2.The Project .................................................................................................................................... 3
2. The Business Model ..................................................................................................................... 3
3. Characterising the Data and Data Processing .............................................................................. 4
4. Data Analysis .....................................................