Student Information

If you are interested in a bachelor’s or master’s theses, please refer to our form. We usually offer a variety of different topics some of which are listed below. Unfortunately, we often have more thesis requests than we have can supervise, so it is possible we might not be able to offer you a topic.

1. Reading up on your topic
2. Short presentation (5min) to the chair on your chosen topic
3. Research and writing (4 months (BA) / 6 months (MA))
4. submission (by email to your supversivor and our secretary)
5. final presentation (20min + questions)

During the whole process we usually meet our supervised students once a week for support and discussion of their progress.

We provide LaTeX templates for your thesis and Ipe presentation templates.

See here for a list of completed bachelor’s / master’s theses.

You are looking for an assistant position related to algorithmic research? Then feel free to contact us. It is likely your assistance is valuable to us. We regularly need teaching assistants (e.g. for designing and correcting exercise sheets) and we also offer research related positions (e.g. as scientific programmer).

Algorithms often perform surprisingly well regarding their running time or solution quality when used on realistic inputs than would be expected from the theoretical worst-case behavior. We try to understand the relevant properties of real-world instances and prove mathematically that these properties actually lead to the good performance. Such insights also help design new algorithms and improve existing ones. Typical techniques used in this task include average case analysis, parameterized algorithms and algorithm engineering.

Refer to Thomas and Marcus.

Energy informatics is a relatively young research topic involving elements from electrical engineering, computer science and economics. We consider optimization problems motivated from real world applications which can be modeled as graphs. Examples are grid extension problems where power lines or technical equipment are added to an existing power grid, and various cable layout problems. Many of these problems are NP-hard, so besides theoretical insights into these problems, we are also interested in efficient heuristics which need to be evaluated in experiments.

Refer to Max Göttlicher and Wendy Yi

Navigating transport networks is an everyday problem. The network is modeled as weighted graph. When using travel time as edge weight, this approach lets us compute fastest routes. Dijkstras well-known algorithm from 1959 finds the desired route, however, transport networks are huge (the graph of Europe alone contains around 45 million road segments) resulting in long query times with classical algorithms. This approach also yields only a single route.

Current research involves supporting interactive query times for large server-based systems with millions of queries every day. This can be achieved by developing speed-up techniques for Dijkstra's algorithm that include pre-processing the network to achieve massive improvements in query time for shortest paths.

Including other modes of transport (car, walking, public transport, …), real time traffic or multi-criteria optimization (e.g. travel time, energy usage, cost, …) require an extended model of the underlying networks. This field offers a wide variety of topics from both theoretical points of view as well as practical implementations with extensive experimental evaluation on real and realistic data sets.

Refer to: Adrian Feilhauer and Michael Zündorf