Evolutionary Equilibrium Detection in Multicriteria Games

Since most real life decisions are multiobjective, multicriteria games offer a more realistic modeling of real-life interactions. Although several equilibrium concepts have been proposed for solving multicriteria games, equilibria detection has not received much attention. Generative relations are proposed to characterize multicriteria equilibria. An evolutionary method based on generative relations is proposed for detecting various multicriteria equilibria: Nash-Pareto, Ideal Nash and Pareto equilibria. Numerical experiments on discrete and continuous games indicate the potential of the proposed approach.