4839. Fleet Sizing under Demand Uncertainty with Contextual Information
Invited abstract in session HE-51: Hub and Service Network Design, stream Hub Location and Network Design.
Thursday, 14:15-15:45Room: HG – Seminar-Raum Geschichte 1 (1st floor, st. 10)
Authors (first author is the speaker)
| 1. | Borzou Rostami
|
| Accounting and Business Analytics, University of Alberta |
Abstract
We study fleet sizing in transportation networks under demand uncertainty through a contextual stochastic optimization framework. The approach incorporates observable contextual factors, such as seasonality and market conditions, to improve demand estimation, while imposing loss-ratio constraints to control expected shortages relative to available capacity. We develop a data-driven solution method combining nonparametric learning and integer programming, establish its asymptotic optimality, and demonstrate its computational tractability in practical settings.
Keywords
- Transportation
- Programming, Integer
- Machine Learning
Status: accepted
Back to the list of papers