In the transportation field, the shift of airline and railway industries toward web-based distribution channels has provided passengers with better access to fare information. This has resulted in passengers becoming more strategic to price. Therefore, a better understanding of passenger choice behavior is required in order to support fare strategies. Methods based on discrete choice (DC) analysis have recently been introduced in revenue management (RM). However, applications of DC models in railway ticket pricing are limited and heterogeneity in choice behavior across different categories of travelers has mostly been ignored. Differences in individual taste are crucial for the RM sector. Additionally, strategic passenger behavior is significant, especially in markets with flexible refund and exchange policy, where ticket cancellation and exchange behavior has been recognized as having major impacts on revenues.

This dissertation examines innovative approaches in discrete choice modeling to support RM systems for intercity passenger railway. The analysis, based on ticket reservation data, contributes to the existing literature in three main aspects. Firstly, this dissertation develops choice models of ticket purchase timing which account for heterogeneity across different categories of passengers. The methodology based on latent class (LC) and mixed logit (ML) model framework offers an alternative approach to demand segmentation without using trip purposes which are not available in the data set used for the analysis.

Secondly, this dissertation develops RM optimization models which use parameters estimated from the choice models and demand functions as key inputs to represent passenger response to RM policy. The approach distinguishes between leisure and business travelers, depending on departure time and day of week. The formulated optimization problem maximizes ticket revenue by simultaneously solving for ticket pricing and seat allocation. Strategies are subjected to capacity constraints determined on the basis of the railway network characteristics.

Finally, this dissertation develops ticket cancellation and exchange model using dynamic discrete choice model (DDCM) framework. The estimated model predicts the timing of ticket cancellations and exchanges 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 tickets and covers the full range of departure time alternatives offered by the railway company.