Abstract
In this thesis, ways of influencing an auctioneer’s revenue are investigated, via auction theory. We have adopted an evolutionary mechanism in designing a method of payment for a winner in a sealed-bid auction. At present, there are two standard methods of payment which may be used by a winner: these are first-price auction and second-price auction. The method of payment proposed in this study lies somewhere between these two traditional methods and is called the -price auction, with the domain of being [0, 1]. In literatures, it was found that without taking the bidders’ behavior into account the price set by the auctioneers would be mis-evaluated. Thus, in my thesis, I have considered the risk preferences of all the bidders in an auction, and have designed a winner’s method of payment.The idea behind this thesis is that by maximizing all of the bidders’ utilities, we may also maximize the auctioneer’s revenue. This kind of problem is a nonlinear optimization problem, so a genetic algorithm has been adopted to solve it. We also implemented our -price auction with real-life cases, and the results are rather promising. Thus, the most important point we wish to demonstrate is that running the -price auction is superior to running either first-price or second-price auctions in a sealed-bid auction situation, which has been shown in theory and in practice.