engleski

Proceedings of the Special Session on Trading Agent Competition @ KES-AMSTA 2012

The Trading Agent Competition (TAC, http://www.sics.se/tac) was established to provide a platform for study of agent behaviour in competitive economic environments. Research teams from around the world (http://www.sics.se/tac/tacparts.php) develop agents for these environments. During annual competitions, they are tested against each other in simulated market environments. Afterwards, competition results can be mined for information on agent behaviours, and their effects on agent performance, market conditions, and the performance and behaviour of competing agents. After each competition, competing agents are made available for offline research. There are several TAC scenarios currently active. The longest active scenario is the Supply Chain Management scenario (TAC SCM, http://www.sics.se/tac/page.php?id=13) where agents compete by buying components, assembling PCs from these components and selling those PCs to customers. Very popular scenario during the last few years is the Ad Auctions scenario (TAC AA, http://aa.tradingagents.org), a sponsored search environment for a simulated advertising scenario where advertisers represent retailers in a simplified home entertainment market. They are placing bids for ads which are displayed to users who are searching home entertainment products by entering various keywords. The newest scenario is the Smart Grid scenario (Power TAC, http://www.powertac.org), where agents act as retail brokers in a local power distribution region, purchasing electrical energy from a wholesale market as well as from local sources, such as homes and businesses with solar panels, and selling power to local customers and into the wholesale market. The goal of this special session is to share and exchange research results about designing agents for TAC scenarios, as well as analysing competition results with a goal of creating policy guidelines. Namely, the special session consists of three papers which describe representatives of University of Zagreb in three different TAC 2012 scenarios – Power TAC, TAC AA and TAC SCM. All the papers in the special session are characterized with multidisciplinary approach, involving the aspects of information science, computer science, artificial intelligence and economics.

Trading Agent Competition ; TAC SCM ; TAC AA ; Power TAC ; KES-AMSTA

nije evidentirano