Komal Muluk: Optimization under Adversarial Uncertainty


An optimization problem under adversarial uncertainty can be essentially formulated as a game between a player and an adversary: The player partially constructs a feasible solution for a given scenario, and then the adversary completes this to a full feasible solution. The goal of the player is to optimize some objective function and the goal of the adversary is to make the player perform as bad as possible. There are various types of adversarial problems. The PhD thesis of Berit Johannes (2011) develops a machinery for deriving hardness results for large classes of the optimization problems with adversarial uncertainty. The thesis only discusses the negative aspects (hardness results) of the area.
The goals of my doctoral project are twofold: On the one hand, the project will derive new negative results, perhaps by extending and generalizing the machinery of Johannes to other families of optimization problems, such as problems in robust optimization. This should lead to new families of hardness and completeness results for the first or the second level of the polynomial hierarchy or for one of the intermediate complexity classes. On the other hand, the goal of the project is to develop positive results for the considered optimization problems. Major emphasis will be put on the investigation of crucial problem parameters, which will be done by applying the tool kit of parameterized complexity. A further goal is the development of fast exact algorithms with decent running times. Finally, the project will identify tractable special cases, for instance by constraining the combinatorics of underlying graph structures, or by imposing additional conditions on underlying cost matrices.