
Research Group for Parameterized Algorithms and Complexity
Parameterized complexity studies algorithmic problems in a multivariate framework that is finer than traditional complexity measures. This more detailed analysis allows us to discover efficient algorithms and reveal structures in algorithmic problems that are hidden from classical approaches. Our research group explores all aspects of parameterized complexity, including algorithms, preprocessing of combinatorial optimization problems (kernelization), and lower bounds (complexity results). Our work involves several problem domains such as graph algorithms, combinatorial optimization problems, constraint satisfaction problems, logic, and the theory of databases.
Head of department

Dániel Marx, Ph.D.
Parent department
Research areas
- parameterized algorithms
- complexity theory
- graph algorithms
- constraint satisfaction
Results
ERC Starting Grant 2012-2016