1 présentation

• 
  10h30 - 12h10

  Non-parametric choice models: estimation and optimization in revenue management

  • Sanjay Dominik Jena, prés., Université du Québec à Montréal

  Given their effective capability of modeling and predicting complex choice behaviour, discrete choice models have been quickly gaining popularity in recent years. In this tutorial, we review recent developments in the area of non-parametric discrete choice models, particularly those based on ranked preference lists. To this end, we will focus on the application of assortment optimization in revenue management. We will discuss both the estimation of such models from transaction data and the subsequent optimization on the obtained choice model in order to derive optimal assortments.