Flexible Decision-Making in Sequential Auctions

Book One

Volume

Proceedings of the AAAI Conference on Artificial Intelligence, 19

SIGART/AAAI Doctoral Consortium

I develop a multi-dimensional sequential auction design space based on Wurman, et al.’s classification, which includes three dimensions: bidding rules, clearing policy, and information revelation policy. The flexible decision-making system is designed to automatically generate strategies for different models with a specification of the parameters. In next section, I present an approach to generate strategies for sequential auctions with discrete strategy spaces. In the following section, I provide two conjectures and discuss an algorithm to compute strategies for sequential auctions with continuous strategy spaces.