Dynamic assortment optimization based on customers’ behavior using transaction data

Abstract

In this paper, we aim at investigating the impact of demand modelling on revenue performance for a shared mobility system. We consider the case where the information associated to the systematic part of the utility function is not available. We use dynamic assortment optimization to estimate utilities. We compare this assumption with the case where the choice probabilities are assumed to be static and homogenously estimated. The results indicate that this additional information helps to improve the service level as well as the revenue performance.