Abstract
In this paper, we introduce an early concept of using multi-objective optimization to study various emerging strategies in evolutionary game theory and show its application in a case study. We aim to analyze the emergent behavior when changing the game’s environment through optimization. The multi-objective approach allows looking at the results of each model evaluation from different points of view. For the realization, we suggest the use of a multi-agent model to compute the outcome of a game. Such a model allows modeling even complex interrelationships and can be used as input to multi-objective optimization algorithms. Finally, we demonstrate a use case by optimizing therapy plans for melanoma through the incorporation of medications into a multi-agent model of concurring cell populations in the tumor micro environment.