Assessment 1 – Group Case Study and Individual Exercise
Your Task
Group Component: Analyse the dataset pertaining to diabetes in a segment of the population that will be provided. Discuss the findings and present insights to the class.
Individual Component: Individually summarise the learnings and suggest additional techniques in a brief report.
Both of these tasks will run in our online class in Week 6.
Assessment Description
Background: We have explored a number of Machine Learning techniques in class. We also observed that that it is possible to abstract these methods using a tool such as Spotfire that has menu driven implementations.
In this assessment you will create a machine learning model that can describe the data and
report of the characteristics of the model and its potential for predicting future cases.
Assessment Instructions
In class: The dataset for this assessment is based on identification of diabetes in a certain population. You will be sent the data set at the beginning of class by your lecturer. As a group:
1. use the appropriate menu driven machine learning method in Spotfire to analyse the dataset or
2. Analyse it utilising a different machine learning implementation.
You will provide an oral presentation of the work in parts A to D during the third hour of the workshop. Are encouraged to use a BI tool to help you explain each section.
The data set will be posted/emailed at the beginning of class in Week 6.
After Class: As an individual, you will take notes in class and then write a 500-word report which summarises the analysis, as well as providing suggestions for further analysis. This component of the assessment is to be handed in by Monday of week 7, at 23.55pm AEDT.
Part A
As a group:
• Load the dataset into Spotfire or your preferred machine learning method
• Create plots that would enable you to assess the characteristics of data
• Create notes for the presentation).
As individuals,
1. Write dot points on this section in your own words (100 words, 5 marks)
Part B
As a group:
Select and appropriate machine learning method, create a model and run the model:
1. What is an appropriate machine learning method to predict diabetes in a patient?
2. What was your output from running the model (accuracy, ROC curve)?
3. Discuss what other methods you could use for prediction?
4. Create notes for presentation
As individuals,
• Write dot points in your own words on the forecasting techniques (200 words, 7 marks).
Part C
As a group, prepare a presentation (10 minutes):
• Present your answers to questions 1-4 from Part B
• You are encouraged to use the BI tool as a method of presentation
• Include key findings
• Highlight methodology.
• All members of the group should be involved in the presentation
As individuals,
• Write dot points in your own words on ways in which the dataset, visualisations and forecasts could be improved (200 words, 8 marks).
NOTE: Individuals will also be given 1 -3 marks for format and grammar in the final report.
