The Automated Negotiation Agents Competition (ANAC2014) was organized [1]. Automated agents can reduce efforts required of people during negotiations and help people for agreeing in negotiations. Therefore, development of automated agents is one of important issues. We developed AgentM based on a heuristic for the ANAC2014 competition. AgentM has three characteristics described as following: AgentM is basically a compromiser agent. The purpose of AgentM’s bidding strategy is to improve own utility and opponent’s utility. AgentM’s accepted strategy is to accept opponent’s bid when it is better than the worst own bid. In this paper, we show AgentM how to compromise, how to improve utility of each other and how to bid.
CITATION STYLE
Niimi, M., & Ito, T. (2016). AgentM. In Studies in Computational Intelligence (Vol. 638, pp. 235–240). Springer Verlag. https://doi.org/10.1007/978-3-319-30307-9_15
Mendeley helps you to discover research relevant for your work.