Scale-Free Networks Emerge out of Micro-Mechanism in an Agent-based Computational Economic Model

The scale-free network structures have been prominently investigated as models of many social and economical networks. However, the topology structures of these complex networks usually lack micro mechanism supporting. In this paper, we propose an agent-based computational economic system - artificial economy market (AEM), in which scale-free transaction networks emerge from local interactions between agents who have the capability of learning through transactions. With a CRA (classifier systems, rules, and actions) decision system, agent can adapt the environment more flexibly. Therefore, due to individual learning and adapting, scale-free transaction networks emerge automatically in our virtual market system. Furthermore, resource propagation takes place on a dynamic scale-free network, and some "pivotal nodes" which have a large number of links emerge in the propagation networks.