Walverine: a Walrasian trading agent

摘要: TAC-02 was the third in a series of Trading Agent Competition events fostering research automating trading strategies by showcasing alternate approaches an open-invitation market game. TAC presents challenging travel-shopping scenario where agents must satisfy client preferences for complementary and substitutable goods interacting through variety types. Michigan's entry, Walverine, bases its decisions on competitive (Walrasian) analysis travel economy. Using this Walrasian model, we construct decision-theoretic formulation optimal bidding problem, which Walverine solves each round good. Walverine's approach, as well several other features overall strategy, are potentially applicable broad class environments.