Here is the step by step instructions for doing the final project.
I am not expecting you to do anything complicated. A simple prediction model would be enough.
Step 1: Find a dataset -here are some sources –[login to view URL] [login to view URL] a dataset that you are most comfortable with–it need not be in healthcare and it need not be from the above two locations either –if you have access to some other data, pl. use that.
Step 2: Develop one or two simple hypotheses with predictors and target variable
Step 3: Use the most appropriate ML technique/sthat you have learned to develop a prediction model.
First, develop a few models –report the accuracy of each model and pick the most accurate one as your final model.
Write a report on the process, methodology and about the model that is the most accurate
What needs to be submitted:Write the above in the format of a short journal paper (2-3 pages–not including references, tables, figures etc)
Introduction –Describe why you are developing this prediction model (e.g. how it will benefit the society). Please add any references if you can find [login to view URL] question and Hypotheses –Write about your predictors and target variable and a simple and clear hypothesis (1 or 2 would be good enough).
Method –Describe the ML method/s you are [login to view URL] all the steps you undertook before developing the model.
Explain why you chose this ML technique over others. Results: Report the accuracy of each model and describe the model that is most accurate.
Appendix: Add tables, figures, R screen shots etc.