Bayesian Hierarchical Choice Modeling Framework for Capturing Heterogeneous Preferences in Engineering System Design

Abstract

Choice models play a critical role in enterprise-driven design by providing a link between engineering design attributes and customer preferences. In our previous work, we introduced the hierarchical choice modeling approach to address the special needs in complex engineering system design. The hierarchical choice modeling approach utilizes multiple model levels to create a link between qualitative attributes considered by consumers when selecting a product and quantitative attributes used for engineering design. In this work, the approach is expanded to the Bayesian Hierarchical Choice Modeling (BHCM) framework, estimated using an All-at-Once (AAO) solution procedure. This new framework addresses the shortcomings of the previous method while providing a highly flexible modeling framework to address the needs of complex system design. In this framework, both systematic and random consumer heterogeneity is explicitly considered, the ability to combine multiple sources of data for model estimation and updating is significantly expanded, and a method to mitigate error propagated throughout the model hierarchy is developed. In addition to developing the new choice model approach, the importance of including a complete representation of consumer heterogeneity in the model framework is provided. The new modeling framework is validated using several metrics and techniques. The benefits of the BHCM method are demonstrated in the design of an automobile occupant package.