Journées de l'optimisation 2017

HEC Montréal, 8-10 mai 2017

1er Atelier Canadien sur l'optimisation des soins de santé (CHOW)

HEC Montréal, 10-11 mai 2017

JOPT2017

HEC Montréal, 8 — 11 mai 2017

Horaire Auteurs Mon horaire
Cal add eabad1550a3cf3ed9646c36511a21a854fcb401e3247c61aefa77286b00fe402

TB4 Optimisation axée sur les données / Data driven optimization

9 mai 2017 10h30 – 12h10

Salle: Meloche Monnex

Présidée par Sanjay Dominik Jena

3 présentations

  • Cal add eabad1550a3cf3ed9646c36511a21a854fcb401e3247c61aefa77286b00fe402
    10h30 - 10h55

    Preference-Based Customer Segmentation for Assortment Planning

    • Gabrielle Gauthier Melancon, Présentateur, JDA Software

    In the data science team at JDA, we developed a preference-based customer segmentation tool that helps in the assortment planning problem, during which a retailer select what products to put in each store. For that, we analyze products’ attributes to discover customers’ motivation and preferences, and then find patterns and similarities between different purchases to group shoppers together.

  • Cal add eabad1550a3cf3ed9646c36511a21a854fcb401e3247c61aefa77286b00fe402
    10h55 - 11h20

    Data-Driven Distributionally-Robust Facility Location Problems

    • Ahmed Saif, Présentateur, HEC Montreal
    • Erick Delage, HEC Montreal

    We consider two classical facility location problems and show how they can be robustified against the distributional ambiguity that arise when using data. We also show that one can improve the worst-case expected performance by “randomizing” over different subsets of locations and devise an efficient algorithm to identify such solutions.

  • Cal add eabad1550a3cf3ed9646c36511a21a854fcb401e3247c61aefa77286b00fe402
    11h20 - 11h45

    Learning consumer preferences for data-driven large-scale assortment optimization

    • Sanjay Dominik Jena, Présentateur, ESG UQAM
    • Andrea Lodi, Polytechnique Montréal
    • Hugo Palmer, Polytechnique Montréal

    We propose a new representation for rank-based choice models that generalizes classical representations. The model allows for subsets of products on which the consumer does not have a strict preference and enables us to efficiently train the consumer preferences. This is exemplified on large artificial and industrial data sets.

Retour