EURO-Online login
- New to EURO? Create an account
- I forgot my username and/or my password.
- Help with cookies
(important for IE8 users)
305. A Contextual Stochastic Optimization Perspective on Demand Prediction for Decision Making
Invited abstract in session TB-1: Emma Frejinger, stream Keynotes.
Tuesday, 10:30-12:00Room: Sportshallen (building: 101)
Authors (first author is the speaker)
| 1. | Emma Frejinger
|
| Université de Montréal |
Abstract
Decision makers are often faced with problems that are subject to uncertainty. Consider the problem of planning transport services for an upcoming season, determining optimal locations of new infrastructure, or establishing production plans and pricing strategies for a product. In this context, demand uncertainty is challenging to deal with, notably because it is decision-dependent. In this talk, we discuss data and modeling challenges associated with understanding and predicting demand. Focusing on the competitive facility location problem, we describe a methodology to deal with decision-dependent demand uncertainty without making strong distributional assumptions. We also provide a high-level overview of contextual stochastic optimization. Studied in the literature under a variety of names, contextual optimization refers to data-driven approaches to prescribe decisions by exploiting relevant side information. We position demand modeling for decision making in this context and outline future research directions on integrated learning and optimization.
Keywords
- Stochastic Optimization
- Optimization Modeling
- Transportation
Status: accepted
Back to the list of papers