Parity games are a class of games played by two players on labelled graphs. They have important connections to logics and model checking, and also interesting complexity properties: it is strongly suspected that the decision problem for parity games is in P, however, it is only known that it is in the intersection of NP and co-NP.
The aim of this project is to enhance understanding of the algorithms by creating an environment which allows to compare different algorithms (known from the literature) with each other, find common traits, and determine which algorithms perform better on sub-classes of the problem space.
This is a high-risk/high-gain project. To alleviate the risk, this work could include the parallelization of algorithms which in any case would be novel.