Welcome to the White-Box AI Research Group at FAU!

Are you passionate about the world of AI and machine learning, including fields such as Business Analytics, Computer Vision, Natural Language Processing, or Interpretable/Explainable Machine Learning? You always wanted to develop a web application, to give users a glimpse into the black-box? You think that we need to talk to users and design experiments to build better machine learning models? Then you’re at the right spot.

At our chair, we focus on making AI more transparent, trustworthy, and accessible. Our diverse thesis topics allow you to delve into various aspects of AI and machine learning, while our dedicated team of researchers and faculty provide guidance and support throughout your thesis journey.

If you’re ready to embark on an engaging and challenging thesis project, explore our available topics listed below and get in touch with us. We look forward to collaborating and making significant contributions to the field of interpretable machine learning and beyond!


The only prerequisite we have at this point is that you’ve taken at least one of the courses offered by either Prof. Mathias Kraus or Prof. Patrick Zschech. However, while we encourage all students to take one of our courses to better understand our research areas, the prerequisite does not apply to students from the following programs:

  • B.Sc. Wirtschaftsinformatik (Information Systems)
  • M.Sc. Internationale Wirtschaftsinformatik (International Information Systems)
  • B.Sc. Informatik (Computer Science)
  • M.Sc. Informatik (Computer Science)

Additionally, please make sure that you independently follow the guidelines of your study program and the examination office. This means that we will not check whether a certain number of ECTS credits has been earned or whether you have taken modules that may be required for admission to your thesis.


We offer two options for selecting a thesis topic:

  1. Students can propose their own topic.
  2. Students can select one from our topic pool that matches their interests and academic goals.

If you choose a topic that involves collaboration with a company, please keep the following in mind:

  • Students must have a company supervisor in addition to their academic supervisor from our department.
  • The results of the thesis project must be publishable in academic settings. We do not accept any confidentiality or non-disclosure agreements that would prevent publication of the results.
  • The data used in the project must be available to our department, and at least the major results of the project must be publishable.
  • It is important to obtain the company’s approval before registering your thesis to ensure that company data is available so that the thesis can be delivered on time.

We encourage you to carefully review the available topics or propose your own topic that is in the same domain or has some related touchpoints to find a research area that matches your interests and academic goals.

Type of Thesis
Textual Heuristic Explanations of White-Box Models’ (EBM, IGANN, PYGAM) for Decision Making Bachelor Thesis

Master Thesis

Nico Hambauer
Generating Synthetic Benchmark Datasets for Enhanced Model Evaluation Master Thesis Sven Kruschel

Lasse Bohlen

Strengthening the Foundations of Model Interpretability: Conceptual Clarification and Practical Approaches Bachelor Thesis

Master Thesis

Julian Rosenberger
Evaluating the Performance of GAMs for predicting Mortality compared to traditional Scoring Systems Bachelor Thesis

Master Thesis

Lasse Bohlen
Uncovering and Mitigating Discrimination Bias in Datasets: A Comprehensive Analysis Bachelor Thesis

Master Thesis

Julian Rosenberger
Exploring Physician Perspectives on the Clinical Implications of Machine Learning-Generated Shape Plots: A Qualitative Study Focused on Generalized Additive Models Bachelor Thesis

Master Thesis

Lasse Bohlen
Feasibility of Textual Explanations for Enhanced Interpretability: A Study on Transforming GAM Shape Plots into Human-Understandable Narratives Bachelor Thesis

Master Thesis

Julian Rosenberger
Developing Effective Guidelines for Collaboration and Documentation on Small-Scale Team Projects in Interpretable Machine Learning: A Case Study on Making a White-Box Model Accessible on Github Bachelor Thesis

Master Thesis

Nico Hambauer
Enhancing Explainability in Healthcare Analytic: Developing Custom Machine Learning Models for Interpretable Health Predictions Based on Time-Series Data Bachelor Thesis

Master Thesis

Lasse Bohlen
Exploring the Role of Transparency and Personalization in User Trust and Data Sharing: A Comparative Analysis of Human and AI-based Decision-Making Systems Bachelor Thesis

Master Thesis

Julian Rosenberger
Derivation of graph metrics that evaluate the interpretability of GAM models and their shape plots.
Bachelor Thesis

Master Thesis

Sven Kruschel


How to apply for a topic

Before applying, please make sure you meet all the requirements for starting a thesis (see requirements above).

To get started, you will need to send an email to your respective advisor expressing your interest in doing a thesis with them.

Include the following documents in your email:

  • A one-page CV highlighting your academic achievements and any relevant experience.
  • A brief motivational statement explaining your research interests and why you want to work on this particular thesis topic (0.5 -1.5 pages).
  • A transcript showing your grades from recent semesters.

It is important to note that incomplete applications will not be considered, so be sure to include all required documents in your email to your advisor.

If you are interested in proposing your own thesis topic, you will also need to include

  • A short introduction to your idea (0.5 pages).
  • A research question that clearly outlines the focus of your proposed dissertation topic.
  • A proposed method and/or study design that you believe would best address your research question.

To help your advisor better understand your background and determine if you’re a good fit for their research project or your proposed thesis topic, it’s a good idea to send along the materials listed below with your email. If you’re applying with your own topic, please send them to Lasse Bohlen. Before sending, make sure to proofread your email and double-check that you’ve attached all required documents.

General Procedure

  1. Choose or propose a topic and follow “How to apply for a topic”.
  2. Before your thesis is accepted, we will check your application and, if necessary, adapt the topic together with you.
  3. After we have confirmed the topic for you, you can start working on the topic and formulate an exposé. This should have the following structure:
    1. Motivation (1 – 1,5 pages)
      1. Relevance
      2. Problem Statement
    2. Related Work (1 – 1,5 pages)
      1. What does already exist?
      2. What is the research gap that will be addressed?
    3. Research Design (1 – 1,5 pages)
      1. Research Questions/Objectives
      2. Research Method/Approach
    4. Structure of this Work (1 page, Graphical Abstract if suitable)
  4. Send your supervisor your finished exposé. Here you can get feedback on your research plan.
  5. Next, register your thesis topic by submitting the final title to the examination board. We will hand out the registration form for that. Please fill out and send us back the registration form.
  6. Now the working and writing phase begins according to your research question. Meanwhile, you can ask your supervisor for feedback or help with problems.
  7. Depending on the agreement with your supervisor, you will either have to give an interim presentation or a final presentation. Arrange a date for this well in advance.
  8. Submit your thesis no later than the deadline given to you by the Examinations Office. Remember that you must hand in: 1 printed copy to the Examinations Office, 1 printed copy at our department and 1 digital copy on USB or CD.  The thesis must be submitted in a bound form.