Final Project


The final project is your opportunity to explore a topic in depth. These projects must run through the entire data science process (and do not forget that it is a chaotic process). Therefore, students must work together in teams, as projects often work in life outside of university.

The deadline is in week 10 of the course according to the schedule. The exact date and time will be announced on Brightspace in consulation with other teachers so as not to burden students with submissions on the same day.

Final Projects

Choose one of the three projects mentioned below,

  • A : Mobility, Built Environment & Sustainability
  • B : Identifying the Health Vulnerability in a City
  • C : Modelling COVID-19 in India

Some examples of good assignments from last year are listed below - [Final projects will be subjected to a plagiraism check, so please do not copy the hypothesis and analysis of students from last year].

  • A : Ambulance Calls
  • B : COVID
  • C : Cycle Network
  • D : Traffic

Milestones (to keep track of your work)

1. Group Creation and Project Selection

Form groups of 4 students each. Each student needs to be part of a group. If you are going to drop the course, don’t sign up for the final project as that may delay the progress of other students. 

Some suggestions for creating effective groups,

  • Be inclusive of people who are not in The Netherlands (ex. Studying from their home country due to COVID restrictions).
  • Strive for diverse groups. (Machine Learning and AI suffer heavily from bias of individuals and communities. Diversity is crucial in meaningful and effective work.) 
  • The most appealing option is to form a group with friends. However, I urge you to spend the first 5 weeks getting to know your peers (I will facilitate that in class) and then form groups where working with each other seems natural.
  • A group of 3 or 5 students will only be accepted as last resort. Please email me with a good motivation in case you want to form such a group.
  • Please email me if you haven’t found a group. I will make sure you do. 
  • Keep in constant contact with your group.

If you have questions,

We will be meeting with our respective TAs to discuss the final project, questions, problems, and concerns. We will meet between 1100-1300 on Wednesdays during Weeks 8, 9 and 10. Please make sure you organise this meeting with your TAs. 

My suggestion is to spend week 7 on data collection and cleaning, problem scoping and getting to know each other, week 8 on EDA, and developing your models, and week 9 and 10 on visualisations, interpretations, and report. 

If there are any questions about the final projects, please reach out to your TAs directly. 

2. Scope of Work and Preliminary EDA

Address the two most important things first:

  • Project statement. The project goal in the posted project description is not fully formulated or tuned. Based on the project description and references, state a well-defined question that you’ll address in the project.
  • Preliminary EDA. Explain your plans for preliminary data exploration. Please take care when planning, so that team members can work individually on these tasks if need be. Stay in touch with your team and communicate regularly.

3. EDA and Revised Project Statement

Formulate the following for yourself:

  • A description of the data: what type of data are you dealing with? What methods have you used to explore the data (initial explorations, data cleaning and reconciliation, etc)?
  • Visualizations and captions that summarize the noteworthy findings of the EDA.
  • A revised project question based on the insights you gained through EDA.
  • A description of a baseline model that you are going to or in the process of building.

4. Project Report

Submit your final project on Brightspace as a pdf (only one submission per group).

  • Use the project template to write the final report.
  • You will be assessed using a project rubric.
  • You are required to also submit a peer evaluation of each team member on Buddycheck through Brightspace which may or may not affect your grade depending on the reports.