Multi-agent Bidding Mechanism with Contract Log Learning Functionality
This paper addresses the agent-based bidding mechanism under trading actions from supplier sites to demand sites. Bidding includes unit price, amount, and storage cost. This trading environment assumes to be completely competitive, which means an agent cannot detect the competitive agent information. To increase the success rate of bidding, the agent must learn its bidding strategy from the past trading log. Our agent estimates the appropriate bidding price from the past “success bids” and “failure bids” by using statistical analysis. Experimental results shows an agent with such learning functionality increases its rate of “success bids” by 45.6%, compared to the agent without such functionality.
KeywordsNegotiation Process Unit Price Coalition Structure Storage Cost Bidding Strategy
Unable to display preview. Download preview PDF.
- 1.Kraus, S.: Agents contracting tasks in non-collaborative environments. In: Proc. of AAAI 1993, pp. 243–248 (1993)Google Scholar
- 2.Zlotkin, G., Rosenschein, J.S.: Conflict resolution in non-cooperative domain. In: Proc. of AAAI 1990, pp. 100–105 (1990)Google Scholar
- 3.Sandholm, T.W., Lesser, V.R.: Advantages of a Leveled Commitment Contracting Protocol. In: Proc. of AAAI 1996, pp. 126–133 (1996)Google Scholar
- 4.Sandholm, T.W., Sikka, S., Norden, S.: Algorithms for optimizing leveled commitment contracts. In: Proc. of IJCAI 1999, pp. 535–540 (1999)Google Scholar
- 5.Rosenschein, J.S.: Consenting agents: Negotiation mechanisms for multi-agent systems. In: Proc. of IJCAI-1993, pp. 792–779 (1993)Google Scholar
- 8.Akiyoshi, M., Komoda, N.: Asynchronous negotiation protocol in supply chains. In: Proc. of IAWTIC 2005, pp. 65–69 (2005)Google Scholar