Bidding Strategies and Value Estimation for the Dutch Flower Auctions
Speaker
Abstract
The Dutch flower auction (DFA) network is the largest in the world, accounting for nearly two-thirds of the global flower trade. Retailers and wholesalers bid in these simultaneous, sequential, multi-unit auctions, with the goal of acquiring flowers to both fulfill committed orders, as well as to meet speculated future demand. In this talk, we present a stylized model of the DFA problem, and propose a Markov Decision Process to find an optimal set of bids via dynamic programming. We then introduce ongoing work in estimating bidders' underlying flower valuations, given observed bidding behavior from real-world DFA data. |
Contactinformation: |
Dr. Wolf Ketter |