Evolutionary design of desired collective behavior

Akira Namatame

National Defense Academy, 1-10-20, Hashirimizu, Yokosuka, 239-8686, Japan

Agent-based modeling enables us to understand emergent phenomena in socio-economic systems. With a properly calibrated model it becomes possible to explore the range of emergent phenomena made possible by the individual-level rules of behavior and interactions among agents. This approach is often referred as the "forward problem". On the other hand, the inverse problem, which consists of designing the rules to create certain collective patterns is very difficult. By combinging evolutionary learning techniques in agent-based modeling, it is possible to explore the space of emergent collective-level behaviors with a view to designing "desired" ones. I will discuss the approach of an evolutionary design of desired collectives using spatial games. Interest: I am interested in multidisciplinary approaches, either empirical or theoretical, to study of complex socio-economic problems. Agents, heterogeneity and interactions are my key concepts to study of the socio-economic systems research. I have been especially working on simulating and synthesizing emergent phenomena and collective behavior of micro-motivated agents. I am also working on showcase applications of agent-based modeling such as market microstructure design, policy analysis, systemic risk, and financial engineering.