Final Project

In most of the cases, the Master Project is a teamwork project of n students working on a subject. Individual projects can also be considered. Projects will naturally be expected to require an effort proportional to the number of team members (300 hours per student). Students may choose a project from a list of suggested projects (see in More information link) or propose one of their own ideas.

Stages of Development of Master’s Project

  • Form a team of 1-2 students.
  • Choose a prioritized list of projects.
  • The Master’s Project Coordinator, Laura Igual (ligual@ub.edu) will assign you a project, by looking for a (maybe global) maximum of the function that represents your global preferences.
  • Discuss the research questions, goals, approach you intend to take, methodology, data needed and time plan with your project supervisor.
  • Construct a logical outline for the project. Include analysis steps and expected outcomes.
  • Define a clear role for all members of the group. This role will be considered in your personal student assessment.
  • Fill in details of the project + role definition in a short document (max 2 pages).
  • Hand in this document to your supervisors and start the project development!

The final project delivery must include:

  • all code produced in a GitHub (or similar) software repository.
  • a project report (see the Report Template directory in  https://github.com/DataScienceUB/PFM  ).
  • a high level description of the completed project (jupyter notebook and/or blog entry).

See this example: https://github.com/axelbrando/Mixture-Density-Networks-for-distribution-and-uncertainty-estimation

Project Assessment

The quality of the projects will be graded by using three sources of information:

  • The delivered information at the end of the project. We will use a “Project Evaluation Rubric” to assign a mark to the project.
  • The public presentation of your project. Your presentation should be professional enough to give at a technical conference (e.g. organized approach, prepared slides, a short demo or video if appropriate).

More information: https://github.com/DataScienceUB/PFM