Abstract
This paper aims to develop a travel demand prediction model of transit services for the elderly in rural depopulated areas. The developed model predicts individual travel frequency of transit services. A discrete-continuous model is adopted to simultaneously represent not only transit choice probability, but also its travel frequency. A simultaneous estimation procedure for the model is also proposed. Furthermore, a non-compensatory utility function is employed to represent conjunctive choice strategies. Results of model estimation show that the discrete-continuous model can properly predict travel frequency of transit services. It is also found that the model is effective to evaluate the effects of improving levels of transit services and consequently can contribute to more efficient transit service planning.