Mentor
Short description of the topic
The first topic of the project is the algorithmic game theory, which is a modern branch of classical game theory, focusing on algorithmic aspects of modelling the behaviour of participants (players) of certain competitive process (game). The motivation is the study of methods, how to control competitive environment with many players and without a central authority which may impose demands on players. It is possible to attain this goal by designing locally defined games, whose Nash equilibria correspond to globally desirable outcomes. This fact is captured by the low price of anarchy of the game. The goal of the project is to establish and publish new results in algorithmic game theory, mainly by designing games where the Nash equilibria correspond to globally desirable outcomes and by investigating their algorithmic and computational complexity aspects.
The second topic deals with the broad range of combinatorial games, mainly in the subfield of the various cops-and-robber settings. Here the setting is usually defined as a game on graph between two players controlling two sets of pawns, where the typical goal of one player is to capture the pawns of the second player. There is however great variability on the rules. Here the task is to study the computational complexity of questions like who wins or how many pawns are neede to win.
The project is theoretical, with the expected outcome in the form of publications in high-ranking conferences or prestigious journals.
Algorithmic game theory Combinatorial game theory MENTOR’S PAGE
Description of the ideal candidate
The applicants are expected to have a strong background in combinatorics, graph theory, foundations of algorithmic game theory or combinatorial game theory. Additionally, specialization in the following areas is a bonus: graph algorithms, cops-and-robber-type problems, computational complexity or approximation algorithms. The applicant must have obtained a PhD in the relevant area no later than 7 years ago.
Details