Multiagent Decision Making: Personal Estimation and Group Utility

Abstract

In multi-agents systems, different agents cooperate with dissimilar concepts, and each agent represents certain person who wants to achieve the most beneficial result. In several cases, some agents have to organize to be a group, and provide one decision to represent this group’s whole volition, espe-cially in Virtual Enterprise applications. In this paper, we are aimed at providing a mechanism to multi-agents decision-making procedure with no partiality of each participating agent in the group. In the first part, we use Fuzzy AHP to as-sist each agent to evaluate a problem. In the second step, we seek to elicit the co-operation level from each agent’s inner world, and each agent will be awarded to a comparative importance according to the cooperation level. Further, the group decision can be easily made up. Finally, we propose a methodology to adapt group’s preference functions. It can make all the participating agents have the chances to pick up the most favorite choice after several rounds of decision-making process.