For Students

I co-supervise special courses, synthesis projects, and BSc/MSc theses at DTU, in collaboration with BrainCapture. This page describes how a project with me gets started and what you can expect along the way — rather than a fixed catalogue of topics.

What we work on

Projects sit somewhere in the space between machine learning research and clinical EEG. Broad directions include:

The concrete question is something we shape together. Bring your own idea if you have one — the best projects usually start from something the student genuinely wants to know.

How we get started

  1. Get in touch early

    Reach out well before the project period starts — ideally 1–2 months ahead. Popular periods fill up, and good scoping takes a few conversations.

  2. Send a short introduction

    A few paragraphs is enough. See the checklist below for what helps me answer quickly.

  3. An informal first meeting

    We talk through your interests, your background, and which of the directions above could fit. No preparation needed beyond having skimmed a paper or two.

  4. Scope the project together

    You write a one-page proposal: research question, data, method, rough milestones, and what "done" looks like. We iterate on it until it is realistic for the number of ECTS you have.

  5. Formalities

    We register the project at DTU with the appropriate main supervisor, sort out any agreements needed for data or company collaboration, and get you access to data and compute.

  6. Run the project

    Weekly or biweekly one-on-ones, plus the option to join the group's research meetings. You are welcome to sit with the team rather than work alone.

  7. Write up — and, where it fits, publish

    Reports are written in the templates below. Strong projects have led to workshop and conference papers, and I am happy to push in that direction when the results support it.

What to include in your first email

Write to wls@braincapture.dk. If you do not hear back within a week, do send a reminder — it is not a no.

Report templates

Start from one of these rather than a blank document. They carry the structure I expect in a report, along with sensible defaults for figures, references, and formatting.

LaTeX template

The conventional choice, and the right one if you may later turn the report into a paper.

Typst template

A modern alternative with much faster compilation. Recommended if you have not committed to LaTeX already.

Expectations, both ways

What I expect from you

Independence and initiative — you drive the project, I help you steer. Come to meetings with something concrete: a result, a plot, or a specific question. Keep your code in version control from day one, and say so early when something is not working.

What you can expect from me

Regular meetings and quick feedback on drafts. Access to real clinical EEG data, GPU compute, and the models we work on. Honest guidance on what is worth pursuing, and support in turning good results into a publication.

← Back to Home