An agent-based taxonomy of adaptation in computional economics

Adaptation, learning and evolution play an important role for the analysis of financial markets in agent-based computational economics. Studying agent adaptation in financial markets can be very informative for understanding the internal workings of the market processes. Conversely, studying (financial) markets modeled as systems of a large number of learning and adapting agents can provide valuable understanding of adaptation and the design of adaptive systems. We propose to investigate this symbiotic relation between agent-based computational economics and adaptive systems by using smart adaptive systems. A starting point for the investigation of smart adaptive agents in computational economics must be the investigation of a framework for adaptation in agent-based systems. A hierarchical taxonomy of adaptation in agent-based systems for computational economics is proposed in this paper. The classification proposed introduces a hierarchy of adaptation schemes, where each level corresponds to the modification of specific components of a generic agent.