Abstract
Conventional factor analysis is difficult to satisfactorily represent fuzziness in subjective evaluation data. This paper therefore attempts to tackle this issue by applying a factor analysis based on fuzzy theory. Moreover, a new discrete choice model (called hybrid discrete choice model) is proposed by incorporating fuzzy latent factors calculated from the above-mentioned factor analysis. Reflecting the influence of fuzziness, utility functions of hybrid discrete choice model are defined as those having interval values. An empirical study, using stated preference data for choice of travel mode under information provision, confirmed the effectiveness of factor analysis based on fuzzy theory and the hybrid discrete choice model.
