Abstract: Agent-oriented models (АОМ) are well-known and are widely used in imitational modeling of economic processes. АОМ allow researchers to take their multicomponent iterative nature into consideration and to avoid difficulties, which may stem from analytical task setting. The article suggests the use Gilboa-Schmeidler? agents, which а) have a limited rationality of decision-making process which accompanies resource exchanges; b) learn, while striking deals aimed at exchange of these resources. This is what makes them different from traditional АОМ, where learning and relearning is not taken into account. The peculiar feature of the agent interaction environment is the possibility of redistribution of a part of resources through the taxation of deals. The influence of institutions on deal successfulness is also taken into account.. The functioning of the suggested АОМ is demonstrated through the use of examples, which admit to simple economic interpretation.

Keywords: agent oriented model, self-learning agents, partial rationality

ACM Classification Keywords: I.2.m Miscellaneous

Link:

IMITATIONAL MODELING OF BEHAVIOR OF LEARNING ECONOMIC AGENTS

Sergey Maruev, Dmitry Stefanovskiy

http://foibg.com/ibs_isc/ibs-24/ibs-24-p07.pdf