Latent class model (LCM) is an analytical technique which is applied to establish model with latent variables. This paper first analyzed the principles of the latent class model, the calibration method of parameters and the evaluation indices after introducing the concepts of latent variables and categorical variables. Then, combined with the number of explicit variables and the number of latent classes which were achieved from the relative fit index, a latent class model of travel mode choice was formulated. Moreover, based on the personal trip survey data of Beijing in 2005, this model was calibrated using Mplus software. At the same time, the proportion of explicit variables in the latent variables was obtained. The calibrated parameters were thoroughly discussed, as well as the characteristics of latent classes. The results show that the latent class model provides a more powerful tool to analyze categorical data for the research of travel mode choice problem.
Study on Travel Mode Choice Behavior Using Latent Class Model
10th Asia Pacific Transportation Development Conference ; 2014 ; Beijing, China
2014-05-05
Conference paper
Electronic Resource
English
Travel Mode Choice Based on Latent Variable Enriched Discrete Choice Model
British Library Conference Proceedings | 2009
|