Master and Bachelor Thesis Supervision (public landing page)

Finding a supervisor

Interested students can contact us with the thesis application form, a curriculum vitae/resume, and an optional cover letter. Please provide evidence for relevant skills. Applicants require significant experience in machine learning, e.g., as acquired through the courses offered by our group (passed with the grade “good” or better) or some equivalent. This typically includes a deep conceptual understanding of machine learning and profound programming experience. The necessary skills may vary depending on the topic, e.g., purely theoretical theses require more mathematical than programming skills.

There are two complementary ways to find a supervisor:

1. Top-down: write a mail to jobs@ml.tu-berlin.de. We will match you with suitable internally advertised topics if possible.

2. Unsolicited: you can contact the researchers below if they work in a research area that suits you.

Many theses are connected to ongoing research in our group. However, it is also possible for students to suggest own topics/ideas. Our group tries to supervise as many students as possible, but we typically do not have the capacity to supervise every interested student.

Researchers:

• ML for Quantum Chemistry: Stefan Chmiela (stefan@chmiela.com)

• Kernel Learning: Stefan Chmiela (stefan@chmiela.com)

• Explainable AI: Grégoire Montavon (gregoire.montavon@tu-berlin.de)

• Anomaly detection/Unsupervised learning: Robert Vandermeulen (vandermeulen@tu-berlin.de)

• Probabilistic modeling and inference: Shinichi Nakajima (nakajima@tu-berlin.de)

• Time series analysis and signal processing: Andreas Ziehe (andreas.ziehe@tu-berlin.de)

• Multimodal signal acquisition and analysis - neurotechnology + biomedical sensing: Alexander von Lühmann (vonluehmann@tu-berlin.de)

• ML for Security & Privacy: Daniel Arp (d.arp@tu-berlin.de)

• Explainable AI & Natural Language Processing: Oliver Eberle (oliver.eberle@tu-berlin.de)

• Learning with multiple modalities: Jannik Wolff (jannik.wolff@tu-berlin.de)

• Robustness against spurious correlations in DNNs, neural cellular automata: Lorenz Linhardt (l.linhardt@campus.tu-berlin.de)

• Combining Explainable AI & Deep Generative Models like Diffusion Models, Large Language Models, GANs, VAEs and Normalizing Flows in different modalities: Sidney Bender (s.bender@tu-berlin.de)

• ML for Quantum Chemistry: Jonas Lederer (jonas.lederer@tu-berlin.de)

You can find further information about our group members on our group's publication list and the BIFOLD website.

Publicly advertised open topics

We advertise most topics internally (link to our internal wiki) because they are related to ongoing and unpublished research.

After having found a supervisor

Students prepare a proposal that includes

• the research question and its context/motivation,
• related work,
• preliminary methodological and/or experimental results,
• and formalities such as the number of ECTS credits and the writing time as specified in the student’s examination regulations.

The supervisor can help the student with writing the proposal. Students can register the thesis with the examination office after Prof. Müller approves the proposal. We encourage students not to underestimate the time required for writing the proposal. Furthermore, consider that we may require some time to review the proposal. Therefore, it is helpful to contact our group early.