Intelligent trading systems: a multi-agent hybrid architecture

We propose an architecture for a multi-agent hybrid intelligent system that can trade in multiple markets. We start by identifying the requirements of an intelligent trading system. We then analyze how intelligent agents can be utilized to construct an intelligent trading system and overcome problems with current approaches. We describe the architecture of the Quo Vadis system (H.-E. Eriksson and M. Penker, 1997) and conclude by showing how the proposed architecture can integrate multiple methodologies for data mining, knowledge discovery and soft computing into a single framework.