Education

Available projects for BSc and MSc theses

Below you find a list of possible topics for BSc and MSc theses in the area of pattern recognition and related fields. Many more projects in the research domain of pattern recognition and machine learning are possible. Please do not hesitate to contact us if you have an interesting idea or want to work something out together. We would love to hear about it – please contact a team member to discuss your idea.

In addition, we also offer topics for BSc theses in the area of human-computer interaction. If you are interested in this particular field, please contact us as well.

The detailed scope of the projects can be discussed and thoroughly defined with the supervisor. That is, the offered projects can be carried out as MSc or BSc thesis.  

For BSc theses a sample template is available. For MSc theses, a sample template is available on the JMCS website.

The document “Formal Requirements and Procedure” describes in detail the formal requirements for your work (such as language or length), the recommended tools, and the entire process from proposal to submission and grading. The document “Academic Writing” provides guidelines for scientific writing and assistance in structuring a thesis. The document “Information for the Pattern Recognition Seminar” contains general information about attending the seminar in our group (attendance is mandatory for students writing their thesis in our group). The document “Giving a Presentation – Checklist for the PRG Seminar“  provides guidelines for giving a talk in our seminar. 

Title	Contact
Analyzing the Impact of Context Length for Large Language Models	corina.masanti@unibe.ch
Extracting Real-Word Errors from Public Datasets for Improved Detection Models	corina.masanti@unibe.ch
Enhancing Gender Neutrality in Text using Large Language Models	corina.masanti@unibe.ch
Learning Graph Edit Distance via Reinforcement Learning	anthony.gillioz@unibe.ch
Deep Metric Learning on Graphs	anthony.gillioz@unibe.ch
Implementing and Evaluating Kolmogorov-Arnold Networks in Graph-Based Contexts	anthony.gillioz@unibe.ch
Efficient Matrix Partitioning Strategies for Accelerated Computation	kalvin.dobler@unibe.ch
Oversmoothing Degradation Analysis on Graph Neural Networks	francesco.leonardi@unibe.ch
Solving Real World Problems with Pattern Classification	kaspar.riesen@unibe.ch

Completed Thesis Projects

Courses