This paper presents an approach to automated mechanism design in the domain of double auctions. We describe a novel parameterized space of double auctions, and then introduce an evolutionary search method that searches this space of parameters. The approach evaluates auction mechanisms using the framework of the TAC Market Design Game and relates the performance of the markets in that game to their constituent parts using reinforcement learning. Experiments show that the strongest mechanisms we found using this approach not only win the Market Design Game against known, strong opponents, but also exhibit desirable economic properties when they run in isolation. © 2011 Elsevier B.V. All rights reserved.
Bibliographical noteThis work was started as part of the first author’s doctoral research at the CUNY Graduate Center and continued as part of the first author’s post-doctoral research at the University of Birmingham and the University of Essex.
This is a revised and substantially extended version of a paper presented at the Twelfth International Workshop on Agent-Mediated Electronic Commerce (Niu et al. 2010a), an extended abstract of which was also presented at the Ninth International Joint Conference on Autonomous Agent and Multi-Agent Systems (Niu et al. 2010b).
- Agent-based computational economics
- CAT game
- Double auction
- Mechanism design
- Trading agent competition