Neural Networks Recommender Agent for Determining Starting Price in Multi Unit Sequential Dutch Auctions
Speaker
Abstract
| In this talk we focus on providing neural networks based recommender agent solution for determining starting price values for Multi-unit Sequential Dutch Auctions (MSDA) auctioneers. First, our agent will predict the first bid and the subsequent bid values. Then, the agent will use the prediction results as basis to give dynamic starting price recommendations. Finally, by adjusting the agent's recommendation tolerance, the auctioneers can preview his decision’s consequences to the revenue and transaction time. |
| Contact information: |
| Meditya Wasesa |
