A data set generation algorithm in combinatorial auctions

Abstract

The generation of realistic data sets in a Combinatorial Auction may be a challenging problem. Well-formed data sets are very useful in the evaluation of algorithms trying to solve the winner determination problem. In this paper a general data set generation scheme is presented, both from an algorithmic and economic point of view. As a case study, a possible auction setting is discussed where the goods on sale are connections between points in space. © 2005 IEEE.