Chaoticity analysis and control of pricing time series in multi-agent artificial demand-supply markets modeled by the genetic programming

Summary form only given. This paper deals with the chaoticity analysis of time series of input pricing on the demand/supply market generated by multi-agent systems and its applications to the control of chaotic behavior of input pricing. The agents' learning processes are modeled by the coevolutionary genetic programming (GP). Five types of agents are introduced and within them a number of agents are assumed to be heterogeneous and to have their own rule for predicting the future pricing. Also we assume agents with random behavior.