For the next two days we will work with the entire end to end pipeline of data science through a case study. We have touched on aspects of this throughout the course but have not yet put all the pieces together.
Topics included in this case study include:
- Classification modeling.
- Programming Practice: Handing off models.
- Teamwork.
- Web applications.
- Website hosting with AWS
- Deploying a DS application.
- Data visualization.
- Results presentation.
- Day 1: Project scoping, Team direction, Model building
- Day 2: Web app and deployment
We will want two deliverables from you for this project:
- A dashboard for investigators to use which helps them identify new events that are worthy of investigation for fraud. This will pull in new data regularly, and update a useful display for the investigation team. How you wish to lay this out is up to you.
- A ten-minute presentation on your process and results.
- Overview: gives a detailed overview of the project. Included are suggestions for how you can organize your team, though this is not binding, and you are free to deviate.
- Building your model: notes on how to get started with the dataset and how to save your model once you've trained it.