Summary
International Symposium on Nonlinear Theory and its Applications
2005
Session Number:1-2-5
Session:
Number:1-2-5-1
Modeling of Agent-based Artificial Auction Markets based on the Genetic Programming and its Applications
Yoshikazu IKEDA, Shozo TOKINAGA,
pp.186-189
Publication Date:2005/10/18
Online ISSN:2188-5079
Summary:
Many real auctions involve complicated trader such as asymmetric bidder, then theoretical analysis becomes very hard. In this paper, we show the agent-based simulation of artificial auction markets by using the Genetic Programming (GP) and its applications. By assuming multi-agents as bidders who learn from past experiences based on the GP, we can analyze the capability to learn successful auctions by agents, and the change of profit of agents in various conditions of auctions. Considering two types of auctions, we can apply the same GP procedure to model learning of agents. In the simulation studies, the effects of parameters such as the private evaluation function are discussed.