Skip to content

Latest commit

 

History

History
39 lines (30 loc) · 2.78 KB

final_project.md

File metadata and controls

39 lines (30 loc) · 2.78 KB

Final Project

Dates

  • Final project proposal due on April 1st
    • The writeup for this can be very short (e.g. 1/2 page or whatever), just so we can all agree on whether it is acceptable. It will not be graded. If we think that the proposal is inappropriate, or want some clarifications, we will tell you that week.
    • If you decide to work in pairs, then you should say something about how work will be divided and coordinated in the project proposal.
  • The final project is due April 25th

If you wish to submit a draft at least 1 week prior to the due date, we will get back to you within 3 days with feedback so you can make some changes.

Requirements

We will accept a wide range of proposals on many different topics. The key requirements are that:

  • Something that involves data and visualization
  • Practice of coding skills with Python (or another language if appropriate and approved by us)
  • If you can relate this to an Honours thesis/ECON490/etc. that is great. But you cannot simply submit your thesis to us. For example, you could take your thesis and do a nice visualization notebook to go with it.
  • Solo work is recommended while a group of two is also okay. For the latter case, grading policy changes accordingly as we expect more than for a single-person project.

See previous final projects

Size of Project and Grading Criteria

You should target roughly a similar length as the amount of code in one of your later problem sets (e.g. 3-5 "pages" of Jupyter notebooks). But having a large notebook is not needed if you are doing something interesting, as you will not be graded solely on the quantity of code.

If you are basing your project on an existing thesis/etc. then we will expect a higher quality of project since you would have a good starting point.

There is no maximum limit to the length of notebooks, but you will not be awarded or penalized for a long notebook.

Grading

There are many ways to do well on this project, for example, you could do well with alternative strategies:

  • If you build a new dataset, it could compensate for less interesting analysis and visualization
  • If you do a highly interactive visualization, it could compensate for using a preexisting dataset
  • If you use some of the new machine learning techniques you learned in the class, you can get by with less interesting datasets
  • Taking the time to make the project public (e.g.a Github repository)

The main criteria for grading will be:

  • Originality
  • Showing one of more examples of techniques learned in the latter part of the course
  • Innovativeness of the idea (although this is not necessary)
  • Clear communication of a lesson/message from the data. Emphasis on clear visualization rather than text.
  • Code clarity and quality