Algorithmen & Komplexität

Informatik 1

Arbeitsgruppe

Forschung

Lehre

Diverses

Kontakt

New Paper in Artificial Intelligence 27 Mar 2026

Opinion dynamics model the spread of opinions in social networks. Usually, they capture local trade-offs for agents who adjust their expressed opinions based on internal convictions and peer pressure in their neighborhood. In our paper Opinion Dynamics with Median Aggregation (by Petra Berenbrink, Martin Hoefer, Dominik Kaaser, Marten Maack, Malin Rau, Lisa Wilhelmi) we study a game-theoretic model for opinion dynamics in combination with a global voting mechanism. Apart from the local neighborhood, agents also take the result of a global vote into account. For the important median voting rule, we analyze existence, structure, and computational complexity of equilibrium outcomes. Moreover, we analyze the impact of changes in the susceptability to the voting outcome. The paper has been accepted for publication in Artificial Intelligence, the international top-journal in AI.

Dissertation Defense 17 Mar 2026

Tim Koglin successfully defended his thesis entitled Algorithms for Information and Contract Design in Multi-Agent Network Settings. Congratulations to Dr. Koglin!

New Paper in ACM Transactions on Computation Theory 18 Feb 2026

Trading claims of a bank in distress is described in, e.g., the U.S. and E.U. banking regulations. We formalize this notion in the context of financial networks and study its algorithmic properties. Our paper "Algorithms for Claims Trading" (by Martin Hoefer, Carmine Ventre, Lisa Wilhelmi) characterizes structural properties, provides efficient (approximation) algorithms to compute optimal claims trades, and shows complementing computational hardness results in several variants. Our results reveal an interesting dichotomy -- optimal trading of claims with a common creditor is algorithmically much easier than trading claims with a common debtor. The paper has been accepted for publication in ACM Transactions on Computation Theory.

New Paper in Theoretical Computer Science 12 Feb 2026

In our paper "The Complexity of Blocking All Solutions" (by Christoph Grüne, Lasse Wulf) we consider a general problem of blocking all solutions of some given combinatorial problem with only few elements. This includes, for example, the problem of destroying all Hamiltonian cycles of a given graph by forbidding only few edges; or the problem of destroying all maximum cliques of a given graph by forbidding only few vertices. For a large class of these problems we show that they are Σ₂^p-complete. Thus, they are substantially more difficult (under standard assumptions in complexity theory) than the underlying base problems (e.g., deciding existence of a Hamiltonian path or finding a maximum clique). The paper has been accepted for publication in Theoretical Computer Science.

Dissertation Defense 9 Feb 2025

Our former group member Komal Muluk successfully defended her thesis entitled Computational Complexity of Graph Orientation, Bilevel and Robust Optimization Problems. Congratulations to Dr. Muluk!