This paper presents an approach to develop bidding agents that participate in multiple alternative auctions, with the goal of obtaining an item with a given probability. The approach consists of a prediction method and a planning algorithm. The prediction method exploits the history of past auctions in order to build probability functions capturing the belief that a bid of a given price may win a given auction. The planning algorithm computes a price, such that by sequentially bidding in a subset of the relevant auctions, the agent can obtain the item at that price with the desired probability. The approach addresses the case where the auctions are for substitutive items with different values. Experimental results show that the approach increases the payoff of their users and the welfare of the market.
展开▼
