Mutual Rationalizability in Vector-Payoff Games

摘要

This paper deals with vector-payoff games, which are also known as Multi-Objective Games (MOGs), multi-payoff games and multi-criteria games. Such game models assume that each of the players does not necessarily consider only a scalar payoff, but rather takes into account the possibility of self-conflicting objectives. In particular, this paper focusses on static non-cooperative zero-sum MOGs in which each of the players is undecided about the objective preferences, but wishes to reveal tradeoff information to support strategy selection. The main contribution of this paper is the introduction of a novel solution concept to MOGs, which is termed here as Multi-Payoff Mutual-Rationalizability (MPMR). In addition, this paper provides a discussion on the development of co-evolutionary algorithms for solving real-life MOGs using the proposed solution concept.