Heterogeneity and Dynamics in Demand Model for Revenue Management with an Application to Intercity Railway Pricing

Methods based on choice behavior have recently been introduced in revenue management to specify the probability of purchasing products as a function of the set of available fares. However, applications of discrete choice models to railway ticket pricing are limited and heterogeneity in choice behavior across different categories of travelers has mostly been ignored. In this presentation we propose choice models of ticket purchase timing decisions which account for heterogeneity across different categories of passengers. The methodology is based on latent class and both parametric and non-parametric mixed logit models. The optimization model developed integrates parameters estimated from the choice models and demand functions to evaluate passenger response to revenue management policy. The second part of the talk presents methods aiming at capturing dynamics in choice behavior. The framework proposed predicts the timing in which ticket cancellation and exchange occurs in response to trip schedule uncertainty, fare, and refund/exchange policy of the railway service. The model is able to predict new departure times of the exchanged ticket and covers the full range of departure time alternatives offered by the railway company.

Biography
Cinzia Cirillo is an Associate Professor at the University of Maryland, Department of Civil and Environmental Engineering. She is interested in both theoretical developments and practical applications of discrete choice models. She has extensive multidisciplinary and international collaborations. She is currently spending six month sabbatical leave at TUDelft under a NWO project.