Hybrid Choice Models with Logit Kernel: Applicability to Large Scale Models1

Abstract

The conceptual framework of the hybrid choice model (HCM) possesses many ingredients to enhance the behavioral representation of the choice process and therefore addresses the problem raised. As a direct result, the choice model specification is improved and it gains in predictive power. The application of HCMs have been limited to small scale models with two or three alternatives. The paper focuses on the major issue that arises from applying hybrid choice models with logit kernel to large scale problems. It concerns probability simulation of large dimensional integrals that arise from the inclusion of numerous attitudes and perceptions in models with large sets of potentially interrelated choices. The paper suggests simulation driven Bayesian and classical approaches to the econometric estimation of models with large number of dimension and flexible choice model formulations. Econometrically, the estimation of hybrid choice models can be extremely involved, but the progress in computer technologies now permits addressing those complicated problems.