🚀 Before starting

🔍 Finding a project

The first step in writing a thesis is to find a project to work on. The four most common ways are:

Find a project at a company, maybe where you are working as a student assistant
Find a project through https://dtu.jobteaser.com
Propose a project yourself
Contact a professor or researcher at DTU for project ideas

Regardless of what way you find a project, it is important that you find a project that you are interested in! You will be working on the project for 5+ months and burnout is a real thing. So, even if you are not in charge of formulating the initial project idea, you can still influence the project during the process to make it more interesting for you. Remember, in the end it is your thesis 🎓

When considering project, think about what parts of computer science that you find most interesting. Very broadly, a thesis within computer science can be divided into four categories(1)

Yes, this is a brute simplification of the field of computer science, but it is a good way to think about what you find interesting and in my experience is still fairly covering.

Theory: This is the mathematical and theoretical part of computer science. It is about understanding the fundamental limits of computation and algorithms. If you like math and theory, this is the place for you.
Software: This is about understanding and working with software. It is about understanding how to design, build and maintain software. If you like programming and software development, this is the place for you.
Data: This is about understanding and working with data. It is about understanding how to store, process and analyze data. If you like working with data, this is the place for you.
Modelling: This is about understanding and working with complex systems. It is about understanding how to model and simulate systems. If you like understanding complex systems, this is the place for you.

You will most likely be working with a combination of these four categories, but it is a good idea to think about what you find most interesting. For example, if you are doing a project in the field of machine learning, you will most likely be working with data and modelling as the primary focus. That said, you can put more focus on software by implementing not just a model but a hole ML system and in the same way you can put more focus on theory by understanding the theoretical limits of the model you are using.

🧙‍♂️ Finding a supervisor

Regardless of how you find a project, you need a supervisor from DTU to register the project and guide you through the project (yes, also for company projects). The DTU rules state that:

Main supervisor

The main supervisor for a master's thesis must

Be an employee at DTU as a VIP with research obligations and one of the following job categories: assistant professor, associate professor, researcher, senior researcher, senior adviser, or professor,
Hold a permanent position (with the exception of assistant professors and researchers)
Have completed DTU's course 'Supervision of larger projects at DTU' or similar pedagogical competence development at the university level e.g. a course in PhD supervision; for assistant professors, however, it is sufficient that one is in the process of UDTU, and
Have in-depth knowledge of the study conditions at DTU.

So not all employees at DTU can be your main-supervisor. That being said, if you find a supervisor that does not meet the above criteria (PhD student, Postdoc or tenure-track assistant professor) it is possible that they can still supervise you, as long as you find a college that meets the above criteria to be your main-supervisor on paper.

If you are doing a company project, you will also need a co-supervisor from the company. The co-supervisor does not officially hold any responsibilities, but they are very much expected to help you with the practicalities of the project and provide you with the necessary data and information.

👓 Read the rules

Yes, it is super boring to read the rules, but it is important to know what you are getting into and you can often find questions to a lot of the questions you might have. So take your time to read them at least once. The rules may differ a bit from studyline to studyline, so make sure you read the rules for your studyline, which you can find here:

https://student.dtu.dk/en/programme-specifications

Especially, be aware of learning objectives for your thesis, as they are the basis for the assessment of your thesis. They will look something like this:

Learning objectives

A graduate of the MSc programme from DTU:

can identify and reflect on technical scientific issues and understand the interaction between the various components that make up an issue
can, based on a clear academic profile, apply elements of current research at an international level to develop ideas and solve problems
masters technical scientific methodologies, theories, and tools, and can take a holistic view of and delimit a complex, open issue, put it into a broader academic and societal perspective, and, on this basis, propose a variety of * possible actions while considering sustainability
can develop relevant models, systems, technologies, and processes aimed at solving technological problems
can communicate and mediate research-based knowledge both orally and in writing
is familiar with and can seek out leading international research within their specialist area.
can work independently and reflect on own learning, academic development, and specialization
masters technical problem-solving at a high level through cross-disciplinary teamwork, and can work with and manage all phases of a project – including preparation of timetables, design, solution, and documentation

All these learning objectives are fairly meta, but they can interpret as you need to master every aspect of doing a thesis from the technical aspects to the communication and project management.

💾 Registering the project

The project needs to be officially registered by your supervisor to DTU study administration. In practice this should be done about two weeks before you start the work. For your supervisor to register the project they need some information.

Registration information

Title: The overall title for the project
Description: A short (5-10 lines) description of the project, that motivates the project and gives an overview of the work to be done. You do not fix the project description in stone, but it is a good idea to have a clear idea of what you want to do.
Number of ECTS: You can do the thesis for 30, 32.5 or 35 ECTS. The default is 30 ECTS.
Start date: The date you start working on the project. If you start in the spring, you are expected to start either in the first week of january or if you are having a 3 weeks course, then the first week after 3 weeks period have ended. Similarly, if you start in the fall, you are expected to start in the first week of August or the end of August depending on if you have a 3 weeks course in June. That said, you can register on alternative days but you need to provide a reason for it.
If company project please also provide:
- Company name
- Company address
- Company contact person
- Company contact email
- Company contact phone number
Student information: Your student number and if it is a group project, the student number of your group members.
Supervisor information: If you have more than one supervisor, you need to provide the email-initials of the ones that are not registering the thesis.

Note that the title and description should be provided in both Danish and English.

🗣️ Setting expectations

As the final step before starting the project, you should set expectations with your supervisor. This is important to make sure that you are on the same page and that you know what is expected of you. Consider the following questions:

What is the expected outcome of the project?
What grade do you expect to get?
What should the meeting schedule be? Regular vs. ad-hoc meetings? Physical vs. online meetings?
What means of communication should be used? Email, Teams, Slack, etc.?