Essays on Advanced Discrete Choice Applications

Abstract

The research contribution of this thesis regarding discrete choice applications and their integration into choice-based optimization approaches can be summarized as follows. For the modeling tasks of school choice, university major choice and fare class choice in airline revenue management approaches are developed to examine individual choice processes and to predict demand for the available choice alternatives.The applied approaches are connected methodologically by covering advanced discrete choice applications that prove to be superior to more simple and, thus, limited approaches. The articles presented in this thesis capture (i) effects of non-constant demand substitution between choice alternatives for the school choice and fare class choice modeling tasks and (ii) the impact of students’ perceptions and beliefs on their choice of a university major. By further exploiting the modeling task of fare class choice, an approach for choice-based optimization in airline revenue management is developed that allows for the incorporation of a general discrete choice demand model with non-constant substitution patterns. As a result, limitations in optimization models that arose from an integration of deterministic demand figures or simple demand models like MNL can be overcome.