An optimization method for agent's bidding strategy in TAC-SCM game

Abstract

Intelligent agents have been developed for a number of e-commerce applications including supply chain management. In Trading Agent Competition for Supply Chain Management (TAC SCM), several manufacturer agents compete in a reverse auction in order to sell assembled computers to customers. In this paper, we consider an individual manufacturer agent in TAC SCM and we focus on the sales task. Using a dynamic programming method, the manufacturer agent is able to find an optimized bidding strategy to decide whether to bid for each arriving request for quote (RFQ). The experiment results show that this strategy improves the agent's revenue significantly comparing to several other heuristics in the current practice. This approach can also be applied to similar bidding problems in other e-commerce applications. © 2010 Springer-Verlag Berlin Heidelberg.